{"id":2163,"date":"2026-09-08T15:45:35","date_gmt":"2026-09-08T15:45:35","guid":{"rendered":"https:\/\/lean-app.com\/?p=2163"},"modified":"2026-09-08T17:55:59","modified_gmt":"2026-09-08T17:55:59","slug":"perdre-du-poids-avec-chatgpt","status":"publish","type":"post","link":"https:\/\/lean-app.com\/en\/perdre-du-poids-avec-chatgpt\/","title":{"rendered":"Perdre du poids avec ChatGPT, \u00e7a marche ? \u00c9tude : 291 r\u00e9ponses, jusqu&rsquo;\u00e0 1 260 kcal d&rsquo;\u00e9cart"},"content":{"rendered":"<div id=\"lvm-shell\"><style>#lvm-shell{display:contents;--rule:#e6e2dc;--ink:#111;--muted:#6b6b6b;--font-display:-apple-system,\"SF Pro Display\",\"Segoe UI\",Roboto,Arial,sans-serif;font-family:var(--font-display)}#lvm-shell .nav{position:sticky;top:0;z-index:50;background:rgba(255,255,255,.86);backdrop-filter:saturate(180%) blur(14px);-webkit-backdrop-filter:saturate(180%) blur(14px);border-bottom:1px solid var(--rule)}#lvm-shell .nav-row{max-width:1160px;margin:0 auto;display:flex;align-items:center;gap:14px;padding:10px 22px;box-sizing:border-box}#lvm-shell .nav-brand{display:flex;align-items:center;gap:9px;text-decoration:none;color:var(--ink)}#lvm-shell .nav-brand img{width:28px;height:28px;border-radius:7px;object-fit:cover}#lvm-shell .nav-brand span{font-weight:600;font-size:18px;letter-spacing:-.01em}#lvm-shell .nav-spacer{flex:1}#lvm-shell .nav-link{color:var(--muted);text-decoration:none;font-size:14px}#lvm-shell .nav-link:hover{color:var(--ink)}#lvm-shell .nav-stores{display:flex;gap:6px;align-items:center}#lvm-shell .nav-stores a{display:block;line-height:0}#lvm-shell .nav-stores img{height:28px;width:auto;border-radius:5px;transition:transform .15s}#lvm-shell .nav-stores a:hover img{transform:translateY(-1px)}@media (max-width:768px){#lvm-shell .nav-row{padding:8px 18px;gap:8px}#lvm-shell .nav-link{display:none}#lvm-shell .nav-stores img{height:24px}#lvm-shell .nav-stores{gap:4px}}@media (max-width:380px){#lvm-shell .nav-stores img{height:22px}}<\/style><header class=\"nav\"><div class=\"nav-row\"><a class=\"nav-brand\" href=\"https:\/\/lean-app.com\/en\/\" aria-label=\"Lean home\"><img loading=\"lazy\" src=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp\" alt=\"\" width=\"512\" height=\"512\" decoding=\"async\"><span>Lean<\/span><\/a><span class=\"nav-spacer\"><\/span><a class=\"nav-link\" href=\"https:\/\/lean-app.com\/en\/tdee-calculator\/\">TDEE Calculator<\/a><div class=\"nav-stores\"><a href=\"https:\/\/apps.apple.com\/fr\/app\/lean-calorie-ai-podometre\/id6738668646?utm_source=seo&amp;utm_medium=blog&amp;utm_campaign=etude-chatgpt\" target=\"_blank\" rel=\"noopener\" aria-label=\"Download on the App Store\"><img loading=\"lazy\" src=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-appstore-official.webp\" alt=\"App Store\" width=\"413\" height=\"122\" decoding=\"async\" style=\"height:48px;width:auto;border-radius:9px;display:block\"><\/a><a href=\"https:\/\/play.google.com\/store\/apps\/details?id=com.lean.testsqflite&amp;utm_source=seo&amp;utm_medium=blog&amp;utm_campaign=etude-chatgpt\" target=\"_blank\" rel=\"noopener\" aria-label=\"Available on Google Play\"><img loading=\"lazy\" src=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-googleplay-official.webp\" alt=\"Google Play\" width=\"315\" height=\"95\" decoding=\"async\" style=\"height:48px;width:auto;border-radius:9px;display:block\"><\/a><\/div><\/div><\/header><\/div>\n<style>#lean-etude{--lean-sticky-top:52px}<\/style><style>\n#lean-etude {\n  --bg: #F3F0EB;\n  --white: #fff;\n  --ink: #222326;\n  --muted: #66635F;\n  --pink: #FF456C;\n  --hot: #FF2D6E;\n  --violet: #DCD0F4;\n  --night: #191A1D;\n  --line: rgb(34 35 38 \/ 15%);\n  --ease: cubic-bezier(.22,1,.36,1);\n  --pointer-x: 0;\n  --pointer-y: 0;\n  --scene-scroll: 0;\n  position: relative;\n  isolation: isolate;\n  container-type: inline-size; container-name: lean-study;\n  width: 100%; max-width: none; margin: 0; padding: 0;\n  color: var(--ink); background: var(--bg);\n  font-family: -apple-system, BlinkMacSystemFont, \"SF Pro Display\", \"SF Pro Text\", \"Segoe UI\", Roboto, Arial, sans-serif;\n  font-size: 16px; font-weight: 400; line-height: 1.6;\n  text-align: left; overflow-wrap: break-word; -webkit-font-smoothing: antialiased;\n}\n#lean-etude *, #lean-etude *::before, #lean-etude *::after { box-sizing: border-box; }\n#lean-etude [hidden] { display: none !important; }\n#lean-etude ::selection { background: var(--pink); color: var(--ink); }\n#lean-etude h1, #lean-etude h2, #lean-etude h3, #lean-etude p, #lean-etude ul, #lean-etude figure { margin: 0; padding: 0; }\n#lean-etude h1, #lean-etude h2, #lean-etude h3 { font-family: inherit; font-style: normal; text-transform: none; color: inherit; text-wrap: balance; }\n#lean-etude h1 { font-size: clamp(52px,6.2cqi,88px); font-weight: 650; line-height: 1.02; letter-spacing: -.04em; }\n#lean-etude h2 { max-width: 23ch; font-size: clamp(36px,4.6cqi,64px); font-weight: 600; line-height: 1.06; letter-spacing: -.035em; }\n#lean-etude h3 { font-size: 24px; font-weight: 600; line-height: 1.25; letter-spacing: -.025em; }\n#lean-etude p { max-width: 72ch; text-wrap: pretty; }\n#lean-etude strong { font-weight: 650; }\n#lean-etude .le-nowrap { white-space: nowrap; }\n#lean-etude ul { list-style: none; }\n#lean-etude li { margin: 0; padding: 0; }\n#lean-etude a { color: inherit; text-decoration: underline; text-underline-offset: .25em; text-decoration-thickness: 1px; }\n#lean-etude a:hover { text-decoration-thickness: 2px; }\n#lean-etude a, #lean-etude button, #lean-etude summary { -webkit-tap-highlight-color: transparent; }\n#lean-etude a:focus-visible, #lean-etude button:focus-visible, #lean-etude summary:focus-visible { outline: 2px solid currentColor; outline-offset: 5px; }\n#lean-etude button { appearance: none; border: 0; padding: 0; color: inherit; font: inherit; font-weight: 600; text-transform: none; letter-spacing: 0; cursor: pointer; }\n#lean-etude .le-section { width: 100%; margin: 0 auto; padding: 100px max(40px,calc((100% - 1280px)\/2)); scroll-margin-top: 80px; }\n#lean-etude .le-section-heading { margin-bottom: 38px; }\n#lean-etude .le-section-label { display: inline-block; margin-bottom: 20px; font-size: 13px; font-weight: 600; }\n#lean-etude .le-progress { position: sticky; top: var(--lean-sticky-top,0px); height: 2px; margin-bottom: -2px; z-index: 5; pointer-events: none; }\n#lean-etude .le-progress i { display: block; width: 100%; height: 100%; transform: scaleX(0); transform-origin: left; background: var(--pink); }\n#lean-etude .le-masthead { display: flex; align-items: center; gap: 36px; max-width: 1440px; height: 100px; margin: auto; padding: 0 40px; }\n#lean-etude .le-brand { font-size: 39px; font-weight: 750; line-height: 1; letter-spacing: -.04em; text-decoration: none; }\n#lean-etude .le-mast-caption { padding-left: 26px; border-left: 1px solid var(--line); font-size: 12px; }\n#lean-etude .le-mast-link { margin-left: auto; min-height: 44px; display: flex; align-items: center; gap: 18px; font-size: 12px; font-weight: 600; text-decoration: none; }\n#lean-etude .le-ouverture { position: relative; display: grid; grid-template-columns: 1.14fr 1fr; grid-template-rows: auto auto; column-gap: 24px; padding-top: 42px; padding-bottom: 46px; }\n#lean-etude .le-hero-copy { position: relative; align-self: center; z-index: 1; padding: 10px 0 30px; }\n#lean-etude .le-eyebrow { display: flex; align-items: center; gap: 10px; margin-bottom: 30px; font-size: 12px; font-weight: 550; }\n#lean-etude .le-eyebrow::before { content: \"\"; width: 6px; height: 6px; border-radius: 50%; background: var(--ink); }\n#lean-etude .le-title-line, #lean-etude .le-title-accent { display: block; }\n#lean-etude .le-title-accent { color: var(--hot); }\n#lean-etude .le-title-follow { display: block; max-width: 28ch; margin-top: 26px; font-size: 26px; font-weight: 500; letter-spacing: -.025em; line-height: 1.2; }\n#lean-etude .le-deck { max-width: 43ch; margin-top: 25px; font-size: 15px; line-height: 1.7; color: var(--muted); }\n#lean-etude .le-hero-actions { display: flex; align-items: center; gap: 25px; margin-top: 27px; }\n#lean-etude .le-explore { display: inline-flex; align-items: center; justify-content: center; min-height: 48px; padding: 12px 24px; border-radius: 50px; background: var(--ink); color: var(--white); font-size: 13px; font-weight: 550; text-decoration: none; transition: transform .3s var(--ease), background-color .3s; }\n#lean-etude .le-scroll-cue { font-size: 22px; }\n#lean-etude .le-byline { max-width: 55ch; margin-top: 24px; font-size: 10px; line-height: 1.7; color: var(--muted); }\n#lean-etude .le-hero-scene { position: relative; align-self: center; min-width: 0; height: 650px; perspective: 1000px; }\n#lean-etude .le-sculpture { position: absolute; inset: -28px -40px auto; width: calc(100% + 80px); height: 690px; max-width: none; object-fit: contain; display: block; mix-blend-mode: normal; transform: translate3d(calc(var(--pointer-x)*-8px),calc(var(--pointer-y)*-5px - var(--scene-scroll)*16px),0); mask-image: linear-gradient(to bottom,#000 87%,transparent 100%); }\n#lean-etude .le-scene-label { position: absolute; right: 6%; bottom: 25px; font-size: 10px; color: var(--muted); }\n#lean-etude .le-scene-orbit { position: absolute; width: 83%; height: 28%; left: 7%; bottom: 8%; border: 1px solid rgb(34 35 38 \/ 15%); border-radius: 50%; transform: rotate(-16deg); pointer-events: none; }\n#lean-etude .le-answer { position: absolute; width: 173px; padding: 14px 16px 13px; color: var(--ink); background: rgb(255 255 255 \/ 84%); backdrop-filter: blur(10px); border-radius: 9px; box-shadow: 0 4px 18px rgb(34 35 38 \/ 7%); transform: translate3d(calc(var(--pointer-x)*var(--depth,12px)),calc(var(--pointer-y)*var(--depth,12px) - var(--scene-scroll)*var(--lift,0px)),0) rotate(var(--angle,0deg)); }\n#lean-etude .le-answer-name { display: block; font-size: 9px; font-weight: 500; line-height: 1.4; }\n#lean-etude .le-answer strong { display: block; margin-top: 6px; font-size: 15px; font-weight: 650; line-height: 1.2; letter-spacing: -.02em; white-space: nowrap; }\n#lean-etude .le-answer-line { display: block; width: 100%; height: 2px; margin-top: 12px; background: var(--pink); transform-origin: left; }\n#lean-etude .le-answer-0 { --depth: 20px; --angle: -7deg; --lift: 20px; left: -8%; top: 21%; }\n#lean-etude .le-answer-1 { --depth: 10px; --angle: 6deg; --lift: -14px; right: -1%; top: 13%; }\n#lean-etude .le-answer-2 { --depth: 28px; --angle: -3deg; --lift: 36px; right: -3%; top: 44%; }\n#lean-etude .le-answer-3 { --depth: 13px; --angle: -8deg; --lift: 12px; left: -6%; bottom: 16%; }\n#lean-etude .le-answer-4 { --depth: 24px; --angle: 6deg; --lift: -6px; right: 0; bottom: 13%; }\n#lean-etude .le-hero-stats { grid-column: 1\/-1; display: grid; grid-template-columns: 1.35fr repeat(3,1fr); gap: 40px; align-items: start; padding-top: 35px; margin-top: 44px; border-top: 1px solid var(--line); }\n#lean-etude .le-hero-stat > strong { display: block; margin-bottom: 12px; font-size: 44px; line-height: .96; letter-spacing: -.035em; font-variant-numeric: tabular-nums; font-weight: 550; }\n#lean-etude .le-hero-stat > span { display: block; max-width: 25ch; font-size: 12px; line-height: 1.5; color: var(--muted); }\n#lean-etude .le-main-stat > strong { font-size: 65px; color: var(--hot); white-space: nowrap; }\n#lean-etude .le-main-stat > strong span { font-size: 19px; letter-spacing: -.01em; }\n#lean-etude .le-nav { position: sticky; top: calc(var(--lean-sticky-top,0px) + 2px); z-index: 4; display: flex; align-items: center; justify-content: center; gap: 8px; margin: 0; padding: 9px 18px; background: var(--bg); border-bottom: 1px solid var(--line); }\n#lean-etude .le-nav a { display: flex; align-items: center; justify-content: center; min-height: 42px; padding: 0 20px; color: var(--ink); border-radius: 32px; font-size: 12px; font-weight: 550; text-decoration: none;  }\n#lean-etude .le-nav a[aria-current=\"location\"] { background: var(--ink); color: var(--white); }\n#lean-etude .le-resume { background: var(--white); display: grid; grid-template-columns: .65fr 1.4fr; gap: 0 70px; }\n#lean-etude .le-resume .le-section-heading { grid-column: 1; grid-row: 1\/3; }\n#lean-etude .le-resume h2 { max-width: 9ch; }\n#lean-etude .le-summary-list { grid-column: 2; }\n#lean-etude .le-summary-list li { display: grid; grid-template-columns: 30px 1fr; gap: 22px; padding: 22px 0; border-bottom: 1px solid var(--line); }\n#lean-etude .le-summary-list li:first-child { padding-top: 0; }\n#lean-etude .le-summary-list li:last-child { border: 0; }\n#lean-etude .le-summary-num { font-size: 16px; line-height: 1.5; font-weight: 500; color: var(--muted); }\n#lean-etude .le-summary-text { font-size: 17px; line-height: 1.55; }\n#lean-etude .le-summary-list li:first-child strong { display: block; margin-bottom: 15px; font-size: 50px; font-weight: 550; letter-spacing: -.035em; line-height: 1; }\n#lean-etude .le-summary-verdict { grid-column: 2; padding: 25px 0 0 52px; margin-top: 15px; font-size: 15px; line-height: 1.7; }\n#lean-etude .le-summary-verdict > strong { color: var(--ink); }\n#lean-etude .le-methode { display: grid; grid-template-columns: .65fr 1.4fr; gap: 25px 70px; }\n#lean-etude .le-methode .le-section-heading { grid-column: 1; grid-row: 1\/3; }\n#lean-etude .le-methode h2 { font-size: 40px; max-width: 10ch; }\n#lean-etude .le-method-stats { grid-column: 2; display: grid; grid-template-columns: repeat(4,minmax(0,1fr)); gap: 30px; }\n#lean-etude .le-method-stats strong { display: block; margin-bottom: 16px; font-size: 55px; font-weight: 500; line-height: 1; letter-spacing: -.035em; }\n#lean-etude .le-method-stats span { display: block; font-size: 12px; line-height: 1.55; }\n#lean-etude .le-method-note { grid-column: 2; padding-top: 24px; border-top: 1px solid var(--line); font-size: 13px; line-height: 1.75; color: var(--muted); }\n#lean-etude .le-simple { --line: rgb(255 255 255 \/ 17%); background: var(--night); color: #F6F4F1; }\n#lean-etude .le-simple .le-section-heading, #lean-etude .le-complet .le-section-heading { max-width: 1050px; }\n#lean-etude .le-result-intro { max-width: 68ch; margin-bottom: 40px; font-size: 17px; line-height: 1.7; }\n#lean-etude .le-question { display: block; max-width: 64ch; margin: 0 0 15px; padding: 0; font-size: 21px; line-height: 1.55; }\n#lean-etude .le-simple .le-result-intro { color: #C9C6C1; font-size: 13px; }\n#lean-etude .le-simple .le-question { color: #F6F4F1; }\n#lean-etude .le-precise { display: block; margin-top: 12px; font-weight: 600; }\n#lean-etude .le-chart-controls { display: flex; align-items: center; justify-content: space-between; gap: 20px; margin: 0 0 0; padding: 20px 25px; background: rgb(255 255 255 \/ 6%); border-radius: 12px 12px 0 0; }\n#lean-etude .le-profile-buttons { display: grid; grid-template-columns: repeat(6,minmax(48px,1fr)); gap: 8px; }\n#lean-etude .le-profile-button { min-width: 54px; min-height: 44px; padding: 11px 18px; border-radius: 30px; background: transparent; font-size: 13px; transition: background-color .2s, color .2s, transform .25s var(--ease); }\n#lean-etude .le-profile-button[aria-pressed=\"true\"] { background: var(--white); color: var(--ink); }\n#lean-etude .le-complet .le-chart-controls { background: #E7E3DD; }\n#lean-etude .le-complet .le-profile-button[aria-pressed=\"true\"] { background: var(--ink); color: var(--white); }\n#lean-etude .le-all-button { min-height: 44px; padding: 10px; background: transparent; font-size: 12px; text-decoration: underline; text-underline-offset: 4px; }\n#lean-etude .le-all-button[aria-pressed=\"true\"] { text-decoration-thickness: 3px; }\n#lean-etude .le-figures { display: grid; gap: 20px; }\n#lean-etude .le-chart { --line: rgb(34 35 38 \/ 12%); display: grid; grid-template-columns: minmax(200px,.72fr) minmax(0,1.8fr); gap: 0 50px; margin: 0; padding: 40px; color: var(--ink); background: #FBFAF8; border-radius: 0 0 12px 12px; }\n#lean-etude .le-profile-title { grid-column: 1; grid-row: 1; align-self: start; font-size: 19px; font-weight: 500; line-height: 1.5; letter-spacing: -.02em; text-wrap: pretty; }\n#lean-etude .le-profile-code { display: inline-flex; align-items: center; justify-content: center; width: 36px; height: 36px; margin: 0 8px 17px 0; border-radius: 50%; background: #E9E4DD; font-size: 12px; font-weight: 600; }\n#lean-etude .le-profile-description { display: block; }\n#lean-etude .le-chart-body { grid-column: 2; grid-row: 1\/3; min-width: 0; min-height: 360px; padding: 4px 0 5px; }\n#lean-etude .le-axis { position: relative; display: flex; justify-content: space-between; padding-bottom: 13px; margin-bottom: 25px; color: #716C66; font-size: 10px; line-height: 1.2; font-variant-numeric: tabular-nums; }\n#lean-etude .le-axis-marks { position: absolute; display: flex; justify-content: space-between; inset: auto 0 0; height: 7px; border-bottom: 1px solid var(--line); }\n#lean-etude .le-axis-marks i { display: block; width: 1px; height: 6px; background: var(--line); }\n#lean-etude .le-chart-row { display: grid; grid-template-columns: minmax(0,1fr) auto; gap: 13px 8px; margin-bottom: 25px; }\n#lean-etude .le-model { font-size: 13px; font-weight: 500; line-height: 1.2; }\n#lean-etude .le-range-value { grid-column: 2; grid-row: 1; font-size: 13px; line-height: 1.2; font-variant-numeric: tabular-nums; font-weight: 600; white-space: nowrap; }\n#lean-etude .le-track { grid-column: 1\/-1; position: relative; height: 14px; background: #EAE6E0; border-radius: 9px; }\n#lean-etude .le-range { position: absolute; left: var(--start); width: var(--span); height: 100%; border-radius: 9px; background: linear-gradient(180deg,#FF829A 0%,#FF456C 57%,#EF315A 100%); box-shadow: 0 3px 8px rgb(255 69 108 \/ 15%); transform-origin: left; }\n#lean-etude .le-range::before, #lean-etude .le-range::after { content: \"\"; position: absolute; top: 5px; width: 4px; height: 4px; background: rgb(255 255 255 \/ 90%); border-radius: 50%; }\n#lean-etude .le-range::before { left: 5px; }\n#lean-etude .le-range::after { right: 5px; }\n#lean-etude .le-target { position: absolute; top: -5px; left: var(--target); width: 2px; height: 24px; background: #171719; transform: translateX(-1px); }\n#lean-etude .le-chart-caption { grid-column: 1; grid-row: 2; align-self: end; padding-top: 23px; margin-top: 28px; border-top: 1px solid var(--line); color: var(--muted); font-size: 11px; line-height: 1.7; }\n#lean-etude .le-caption-range { color: var(--ink); font-weight: 600; }\n#lean-etude .le-caption-gap { display: block; margin: 14px 0; color: var(--ink); font-size: 29px; font-weight: 600; line-height: 1.2; letter-spacing: -.025em; }\n#lean-etude .le-behavior { margin-top: 55px; }\n#lean-etude .le-behavior-title { max-width: none; margin-bottom: 28px; font-size: 21px; font-weight: 500; line-height: 1.3; }\n#lean-etude .le-behavior-list { display: grid; grid-template-columns: repeat(5,minmax(0,1fr)); gap: 36px; }\n#lean-etude .le-behavior-list li { display: grid; grid-template-rows: 80px auto; position: relative; gap: 18px; }\n#lean-etude .le-behavior-list strong { display: flex; align-items: center; justify-content: center; width: 80px; height: 80px; grid-row: 1; grid-column: 1; font-size: 25px; font-weight: 500; letter-spacing: -.02em; font-variant-numeric: tabular-nums; }\n#lean-etude .le-behavior-list li > span:not(.le-behavior-meter) { grid-row: 2; font-size: 12px; line-height: 1.6; }\n#lean-etude .le-behavior-meter { grid-row: 1; grid-column: 1; width: 80px; height: 80px; border-radius: 50%; background: conic-gradient(var(--pink) var(--percent),rgb(128 128 128 \/ 18%) 0); mask: radial-gradient(farthest-side,transparent calc(100% - 3px),#000 calc(100% - 2.5px)); transform: rotate(-90deg); }\n#lean-etude .le-behavior-meter i { display: none; }\n#lean-etude .le-insight { max-width: 68ch; margin: 45px 0 0 auto; padding-top: 30px; border-top: 1px solid var(--line); font-size: 17px; line-height: 1.7; }\n#lean-etude .le-simple .le-insight { color: #D2CFCA; }\n#lean-etude .le-simple .le-insight strong { color: var(--white); }\n#lean-etude .le-complet { padding-top: 100px; }\n#lean-etude .le-complet .le-chart { background: var(--white); }\n#lean-etude .le-comparison { margin-top: 75px; }\n#lean-etude .le-comparison-title { margin-bottom: 32px; font-size: 30px; }\n#lean-etude .le-comparison-stats { display: grid; grid-template-columns: repeat(3,minmax(0,1fr)); gap: 0; border-top: 1px solid var(--line); border-bottom: 1px solid var(--line); }\n#lean-etude .le-comparison-stats li { padding: 35px 35px 35px 0; }\n#lean-etude .le-comparison-stats li + li { padding-left: 35px; border-left: 1px solid var(--line); }\n#lean-etude .le-comparison-stats strong { display: block; margin-bottom: 22px; font-size: clamp(44px,5.5cqi,74px); font-weight: 500; line-height: .95; letter-spacing: -.035em; white-space: nowrap; }\n#lean-etude .le-comparison-stats small { font-size: 18px; font-weight: 500; letter-spacing: -.01em; }\n#lean-etude .le-comparison-stats span { display: block; max-width: 27ch; font-size: 13px; line-height: 1.6; }\n#lean-etude .le-comparison-stats li:last-child strong { color: var(--hot); }\n#lean-etude .le-suivi { --line: rgb(255 255 255 \/ 17%); color: #F6F3EF; background: var(--night); padding-bottom: 0; }\n#lean-etude .le-suivi h2 { max-width: 27ch; font-size: clamp(35px,5cqi,65px); }\n#lean-etude .le-time-headline { display: block; margin-top: 14px; color: #FF7795; }\n#lean-etude .le-time-intro { max-width: 51ch; margin: 0 0 60px auto; font-size: 17px; line-height: 1.7; color: #D5D1CB; }\n#lean-etude .le-time-stage { --time-progress: 0; display: grid; grid-template-columns: 1fr 1fr; column-gap: 100px; align-items: start; margin-bottom: 75px; }\n#lean-etude .le-time-dial { position: sticky; top: calc(var(--lean-sticky-top,0px) + 110px); padding-bottom: 40px; text-align: center; }\n#lean-etude .le-dial-face { position: relative; display: flex; flex-direction: column; align-items: center; justify-content: center; width: 100%; max-width: 390px; aspect-ratio: 1; margin: 0 auto; border-radius: 50%; background: radial-gradient(circle at 40% 40%,#353035 0%,#222124 55%,#191A1D 72%); }\n#lean-etude .le-dial-face::before { content: \"\"; position: absolute; inset: 0; border-radius: 50%; background: conic-gradient(from -90deg,#FF7795 calc(var(--time-progress)*100%),rgb(255 255 255 \/ 14%) 0); mask: radial-gradient(farthest-side,transparent calc(100% - 2px),#000 calc(100% - 1.5px)); }\n#lean-etude .le-dial-face::after { content: \"\"; position: absolute; inset: 15px; border: 1px solid rgb(255 255 255 \/ 7%); border-radius: 50%; }\n#lean-etude .le-dial-unit { font-size: 13px; }\n#lean-etude .le-dial-day { margin-top: 0; color: #FF7795; font-size: clamp(80px,12cqi,170px); font-weight: 400; line-height: 1.05; letter-spacing: -.04em; font-variant-numeric: tabular-nums; }\n#lean-etude .le-dial-caption { display: block; margin-top: 28px; font-size: 12px; color: #BEBAB4; }\n#lean-etude .le-timeline { padding-top: 0; }\n#lean-etude .le-timeline li { position: relative; min-height: 215px; padding: 25px 0 45px 33px; border-top: 1px solid var(--line); }\n#lean-etude .le-timeline li::before { content: \"\"; position: absolute; top: 30px; left: 0; width: 7px; height: 7px; border-radius: 50%; background: #777477; transition: background-color .3s; }\n#lean-etude .le-timeline li.is-current::before { background: #FF7795; box-shadow: 0 0 16px rgb(255 69 108 \/ 35%); }\n#lean-etude .le-timeline strong { display: block; margin-bottom: 20px; font-size: 17px; font-weight: 500; line-height: 1.3; }\n#lean-etude .le-timeline strong b { font-size: inherit; font-weight: inherit; }\n#lean-etude .le-timeline li > span { display: block; max-width: 34ch; font-size: 21px; line-height: 1.5; letter-spacing: -.02em; }\n#lean-etude .le-timeline li.is-current > span { color: #FFB0C2; }\n#lean-etude .le-checklists { display: grid; grid-template-columns: 1fr 1fr; gap: 85px; padding: 55px 0 80px; border-top: 1px solid var(--line); }\n#lean-etude .le-checklist h3 { max-width: 20ch; margin-bottom: 32px; font-size: 30px; font-weight: 500; }\n#lean-etude .le-checklist ul { display: grid; gap: 19px; }\n#lean-etude .le-checklist li { display: flex; gap: 14px; align-items: baseline; font-size: 14px; line-height: 1.7; color: #D2CFC9; }\n#lean-etude .le-checklist li > span:first-child { flex: 0 0 18px; color: #FF7795; font-size: 14px; }\n#lean-etude .le-checklist-yes li > span:first-child { color: #A9D7C5; }\n#lean-etude .le-verdict { margin: 0; padding: 75px max(40px,calc((100% - 1280px)\/2)); background: #DCD0F4; color: var(--ink); font-size: clamp(26px,3.5cqi,45px); font-weight: 400; letter-spacing: -.03em; line-height: 1.35; text-wrap: pretty; }\n#lean-etude .le-verdict > strong { display: block; max-width: none; margin-bottom: 34px; font-size: 13px; font-weight: 550; line-height: 1.6; letter-spacing: 0; }\n#lean-etude .le-app { display: grid; grid-template-columns: 1fr 1fr; gap: 30px 80px; background: var(--white); }\n#lean-etude .le-app .le-section-heading { margin: 0; }\n#lean-etude .le-app h2 { max-width: 15ch; font-size: 46px; }\n#lean-etude .le-app-intro { grid-column: 1; font-size: 16px; }\n#lean-etude .le-app-panel { grid-column: 2; grid-row: 1\/3; align-self: center; padding: 0 0 0 45px; border-left: 1px solid var(--line); }\n#lean-etude .le-feature-list { display: flex; flex-wrap: wrap; gap: 12px 20px; padding-bottom: 28px; margin-bottom: 30px; border-bottom: 1px solid var(--line); }\n#lean-etude .le-feature { display: inline-flex; align-items: baseline; gap: 8px; font-size: 11px; }\n#lean-etude .le-feature::before { content: \"\"; flex: 0 0 6px; width: 6px; height: 6px; background: #FA256C; border-radius: 50%; }\n#lean-etude .le-feature-1::before { background: #FF8C00; }\n#lean-etude .le-feature-2::before { background: #E84D44; }\n#lean-etude .le-feature-3::before, #lean-etude .le-feature-4::before { background: #4CAF50; }\n#lean-etude .le-feature-4 { flex-basis: 100%; }\n#lean-etude .le-app-panel > strong { display: block; max-width: 22ch; margin-bottom: 18px; font-size: 25px; font-weight: 500; letter-spacing: -.025em; line-height: 1.25; }\n#lean-etude .le-app-panel > span { display: block; font-size: 13px; line-height: 1.75; color: var(--muted); }\n#lean-etude .le-store-links { display: flex; flex-wrap: wrap; gap: 18px; margin-top: 20px; }\n#lean-etude .le-store-links a { display: flex; align-items: center; min-height: 44px; font-size: 12px; font-weight: 600; line-height: 1.4; }\n#lean-etude .le-donnees h2 { max-width: 19ch; }\n#lean-etude .le-data-list { display: grid; grid-template-columns: repeat(2,minmax(0,1fr)); column-gap: 50px; row-gap: 0; margin-bottom: 35px; }\n#lean-etude .le-data-record { padding: 28px 0; border-top: 1px solid var(--line); }\n#lean-etude .le-data-heading { margin-bottom: 17px; font-size: 16px; font-weight: 500; line-height: 1.5; text-wrap: pretty; }\n#lean-etude .le-data-heading strong { font-weight: 700; }\n#lean-etude .le-data-values { font-size: 13px; line-height: 1.9; color: var(--muted); }\n#lean-etude .le-data-values strong { color: var(--ink); font-weight: 600; font-variant-numeric: tabular-nums; }\n#lean-etude .le-data-note { max-width: 85ch; margin: 18px 0; font-size: 13px; line-height: 1.8; color: var(--muted); }\n#lean-etude .le-data-note:last-child { margin-top: 25px; font-size: 11px; }\n#lean-etude .le-disclosure { border-bottom: 1px solid var(--line); }\n#lean-etude .le-disclosure summary { position: relative; display: block; list-style: none; min-height: 68px; padding: 24px 40px 24px 0; font-size: 17px; font-weight: 500; line-height: 1.4; letter-spacing: -.02em; cursor: pointer; }\n#lean-etude .le-disclosure summary::-webkit-details-marker { display: none; }\n#lean-etude .le-disclosure summary::before, #lean-etude .le-disclosure summary::after { content: \"\"; position: absolute; right: 4px; top: 33px; width: 14px; height: 1px; background: currentColor; transition: transform .25s var(--ease); }\n#lean-etude .le-disclosure summary::after { transform: rotate(90deg); }\n#lean-etude .le-disclosure[open] summary::after { transform: rotate(0); }\n#lean-etude .le-disclosure p { max-width: 75ch; margin: 0 0 22px; font-size: 14px; line-height: 1.8; color: var(--muted); }\n#lean-etude .le-disclosure ul { max-width: 80ch; margin: 0 0 24px; padding-left: 18px; list-style: disc; font-size: 13px; line-height: 1.8; }\n#lean-etude .le-disclosure li + li { margin-top: 13px; }\n#lean-etude .le-faq { display: grid; grid-template-columns: .65fr 1.4fr; gap: 0 70px; background: var(--white); }\n#lean-etude .le-faq .le-section-heading { grid-column: 1; grid-row: 1\/8; }\n#lean-etude .le-faq h2 { font-size: 43px; }\n#lean-etude .le-faq .le-disclosure { grid-column: 2; }\n#lean-etude .le-faq .le-disclosure:first-of-type summary { padding-top: 4px; }\n#lean-etude .le-faq .le-disclosure:first-of-type summary::before, #lean-etude .le-faq .le-disclosure:first-of-type summary::after { top: 13px; }\n#lean-etude .le-suite { color: #F6F4F1; background: var(--night); }\n#lean-etude .le-suite h2 { font-size: 35px; }\n#lean-etude .le-related { display: grid; grid-template-columns: repeat(2,minmax(0,1fr)); column-gap: 65px; row-gap: 0; }\n#lean-etude .le-related li { border-top: 1px solid rgb(255 255 255 \/ 18%); }\n#lean-etude .le-related a { position: relative; display: block; height: 100%; padding: 27px 45px 27px 0; font-size: 18px; font-weight: 500; line-height: 1.4; text-decoration: none; }\n#lean-etude .le-related a::after { content: \"\u2197\"; position: absolute; right: 0; top: 26px; font-size: 23px; transition: transform .25s var(--ease); }\n#lean-etude .le-related a span { display: block; margin-top: 14px; font-size: 12px; font-weight: 400; line-height: 1.7; color: #BFBAB5; }\n#lean-etude .le-suivi { padding: 0; }\n#lean-etude .le-suivi-inner { padding: 100px max(40px,calc((100% - 1280px)\/2)) 0; }\n@media (hover:hover) {\n  #lean-etude .le-profile-button:hover:not([aria-pressed=\"true\"]) { background: rgb(128 128 128 \/ 15%); }\n  #lean-etude .le-nav a:hover:not([aria-current]) { background: #E6E0D9; }\n  #lean-etude .le-explore:hover { transform: translateY(-2px); background: #3C3C3F; }\n  #lean-etude .le-related a:hover::after { transform: translate(3px,-3px); }\n}\n#lean-etude button:active, #lean-etude .le-explore:active { transform: scale(.96); }\n@container lean-study (max-width:1100px) {\n  #lean-etude .le-suivi-inner { padding-inline: 32px; }\n  #lean-etude .le-section { padding-inline: 32px; }\n  #lean-etude .le-ouverture { grid-template-columns: 1.05fr 1fr; }\n  #lean-etude h1 { font-size: 65px; }\n  #lean-etude .le-title-follow { font-size: 23px; }\n  #lean-etude .le-hero-scene { height: 585px; }\n  #lean-etude .le-sculpture { inset: 0 -25px auto; width: calc(100% + 50px); height: 570px; }\n  #lean-etude .le-answer { width: 150px; padding: 12px; }\n  #lean-etude .le-answer strong { font-size: 13px; }\n  #lean-etude .le-answer-0 { left: -5%; }\n  #lean-etude .le-answer-2 { right: 0; }\n  #lean-etude .le-answer-3 { left: -2%; }\n  #lean-etude .le-resume, #lean-etude .le-methode, #lean-etude .le-faq { column-gap: 40px; }\n  #lean-etude .le-method-stats { gap: 22px; }\n  #lean-etude .le-chart { gap: 0 30px; padding: 32px; grid-template-columns: minmax(165px,.7fr) minmax(0,1.8fr); }\n  #lean-etude .le-profile-title { font-size: 17px; }\n  #lean-etude .le-caption-gap { font-size: 26px; }\n  #lean-etude .le-behavior-list { gap: 25px; }\n  #lean-etude .le-comparison-stats strong { font-size: 56px; }\n  #lean-etude .le-time-stage { column-gap: 65px; }\n  #lean-etude .le-checklists { gap: 50px; }\n  #lean-etude .le-app { column-gap: 40px; }\n  #lean-etude .le-app-panel { padding-left: 30px; }\n  #lean-etude .le-verdict { padding-inline: 32px; }\n}\n@container lean-study (max-width:800px) {\n  #lean-etude .le-suivi-inner { padding: 74px 28px 0; }\n  #lean-etude .le-masthead { height: 82px; padding-inline: 28px; }\n  #lean-etude .le-section { padding: 74px 28px; }\n  #lean-etude .le-ouverture { padding-top: 28px; padding-bottom: 34px; gap: 0 12px; }\n  #lean-etude h1 { font-size: 51px; }\n  #lean-etude .le-eyebrow { font-size: 10px; margin-bottom: 22px; }\n  #lean-etude .le-title-follow { margin-top: 20px; font-size: 21px; }\n  #lean-etude .le-deck { font-size: 14px; }\n  #lean-etude .le-hero-scene { height: 530px; }\n  #lean-etude .le-sculpture { inset: 0 -20px auto; width: calc(100% + 40px); height: 520px; }\n  #lean-etude .le-answer { width: 133px; padding: 10px; }\n  #lean-etude .le-answer-name { font-size: 8px; }\n  #lean-etude .le-answer strong { font-size: 12px; }\n  #lean-etude .le-answer-0 { left: -4%; top: 19%; }\n  #lean-etude .le-answer-1 { right: 0; top: 6%; }\n  #lean-etude .le-answer-3 { bottom: 25%; left: -3%; }\n  #lean-etude .le-answer-4 { bottom: 12%; }\n  #lean-etude .le-hero-stats { gap: 24px; grid-template-columns: 1.2fr repeat(3,1fr); margin-top: 20px; }\n  #lean-etude .le-hero-stat > strong { font-size: 38px; }\n  #lean-etude .le-main-stat > strong { font-size: 49px; }\n  #lean-etude .le-main-stat > strong span { font-size: 15px; }\n  #lean-etude .le-hero-stat > span { font-size: 11px; }\n  #lean-etude h2 { font-size: 43px; }\n  #lean-etude .le-resume, #lean-etude .le-methode { grid-template-columns: .6fr 1fr; gap: 25px 40px; }\n  #lean-etude .le-summary-text { font-size: 15px; }\n  #lean-etude .le-summary-list li:first-child strong { font-size: 35px; }\n  #lean-etude .le-summary-list li { gap: 14px; grid-template-columns: 20px 1fr; }\n  #lean-etude .le-summary-verdict { padding-left: 34px; font-size: 14px; }\n  #lean-etude .le-method-stats { grid-template-columns: repeat(2,minmax(0,1fr)); gap: 28px; }\n  #lean-etude .le-method-note { font-size: 12px; }\n  #lean-etude .le-chart { gap: 0 24px; padding: 28px; }\n  #lean-etude .le-chart-caption { font-size: 10px; }\n  #lean-etude .le-caption-gap { font-size: 24px; }\n  #lean-etude .le-model, #lean-etude .le-range-value { font-size: 12px; }\n  #lean-etude .le-comparison-stats strong { font-size: 45px; }\n  #lean-etude .le-comparison-stats li { padding-right: 24px; }\n  #lean-etude .le-comparison-stats li + li { padding-left: 24px; }\n  #lean-etude .le-comparison-stats small { font-size: 14px; }\n  #lean-etude .le-time-stage { column-gap: 38px; grid-template-columns: .8fr 1fr; }\n  #lean-etude .le-timeline li > span { font-size: 18px; }\n  #lean-etude .le-app h2 { font-size: 38px; }\n  #lean-etude .le-verdict { padding: 60px 28px; }\n}\n@container lean-study (max-width:640px) {\n  #lean-etude .le-suivi-inner { padding: 60px 22px 0; }\n  #lean-etude .le-masthead { height: 76px; padding-inline: 22px; gap: 18px; }\n  #lean-etude .le-brand { font-size: 33px; }\n  #lean-etude .le-mast-caption { display: none; }\n  #lean-etude .le-mast-link { font-size: 10px; }\n  #lean-etude .le-section { padding: 60px 22px; }\n  #lean-etude .le-ouverture { display: grid; grid-template-columns: minmax(0,1fr); padding-top: 22px; padding-bottom: 30px; }\n  #lean-etude .le-hero-copy { display: contents; }\n  #lean-etude .le-ouverture .le-eyebrow { grid-row: 1; }\n  #lean-etude .le-ouverture h1 { grid-row: 2; }\n  #lean-etude .le-ouverture .le-hero-scene { grid-row: 3; }\n  #lean-etude .le-ouverture .le-deck { grid-row: 4; }\n  #lean-etude .le-ouverture .le-hero-actions { grid-row: 5; }\n  #lean-etude .le-ouverture .le-byline { grid-row: 6; }\n  #lean-etude .le-ouverture .le-hero-stats { grid-row: 7; }\n  #lean-etude h1 { font-size: clamp(40px,10.4cqi,62px); line-height: 1.04; font-weight: 650; }\n  #lean-etude .le-eyebrow { margin-bottom: 22px; font-size: 11px; }\n  #lean-etude .le-title-follow { max-width: 26ch; margin-top: 22px; font-size: 22px; }\n  #lean-etude .le-deck { margin-top: 21px; font-size: 14px; }\n  #lean-etude .le-hero-actions { margin-top: 22px; }\n  #lean-etude .le-explore { min-height: 44px; padding: 11px 21px; font-size: 12px; }\n  #lean-etude .le-byline { max-width: 57ch; margin-top: 20px; font-size: 9px; }\n  #lean-etude .le-hero-scene { height: 465px; margin-top: 17px; }\n  #lean-etude .le-sculpture { inset: -5px 0 auto; width: 100%; height: 470px; }\n  #lean-etude .le-answer { width: 138px; padding: 10px 12px; }\n  #lean-etude .le-answer-name { font-size: 8px; }\n  #lean-etude .le-answer strong { font-size: 12px; }\n  #lean-etude .le-answer-line { margin-top: 9px; }\n  #lean-etude .le-answer-0 { top: 36%; left: 0; }\n  #lean-etude .le-answer-1 { top: 5%; right: 0; }\n  #lean-etude .le-answer-2 { top: 41%; right: 0; }\n  #lean-etude .le-answer-3 { bottom: 19%; left: 0; }\n  #lean-etude .le-answer-4 { bottom: 9%; right: 0; }\n  #lean-etude .le-scene-label { bottom: 0; right: 0; font-size: 9px; }\n  #lean-etude .le-hero-stats { grid-template-columns: repeat(3,minmax(0,1fr)); gap: 30px 24px; margin-top: 35px; padding-top: 25px; }\n  #lean-etude .le-main-stat { grid-column: 1\/-1; display: flex; align-items: center; justify-content: space-between; gap: 20px; padding-bottom: 25px; border-bottom: 1px solid var(--line); }\n  #lean-etude .le-main-stat > strong { flex: 0 0 auto; margin: 0; font-size: 65px; }\n  #lean-etude .le-main-stat > strong span { display: block; margin-top: 8px; font-size: 16px; }\n  #lean-etude .le-main-stat > span { max-width: 19ch; font-size: 13px; }\n  #lean-etude .le-mini-stat > strong { font-size: 35px; }\n  #lean-etude .le-mini-stat > span { font-size: 10px; }\n  #lean-etude .le-nav { justify-content: space-between; gap: 0; padding: 7px 12px; }\n  #lean-etude .le-nav a { min-height: 44px; padding: 0 9px; font-size: 10px; }\n  #lean-etude h2 { font-size: clamp(35px,8.5cqi,48px); }\n  #lean-etude .le-section-label { font-size: 11px; margin-bottom: 18px; }\n  #lean-etude .le-section-heading { margin-bottom: 30px; }\n  #lean-etude .le-resume, #lean-etude .le-methode, #lean-etude .le-faq, #lean-etude .le-app { display: block; }\n  #lean-etude .le-resume h2 { max-width: 16ch; }\n  #lean-etude .le-summary-list li:first-child strong { font-size: 44px; }\n  #lean-etude .le-summary-list li { grid-template-columns: 18px 1fr; gap: 18px; padding-block: 22px; }\n  #lean-etude .le-summary-text { font-size: 16px; }\n  #lean-etude .le-summary-num { font-size: 13px; }\n  #lean-etude .le-summary-verdict { padding-left: 36px; margin-top: 10px; font-size: 14px; }\n  #lean-etude .le-methode h2 { max-width: 15ch; font-size: 36px; }\n  #lean-etude .le-method-stats { margin: 36px 0; grid-template-columns: repeat(4,minmax(0,1fr)); gap: 15px; }\n  #lean-etude .le-method-stats strong { margin-bottom: 12px; font-size: 40px; }\n  #lean-etude .le-method-stats span { font-size: 10px; line-height: 1.6; }\n  #lean-etude .le-method-note { font-size: 12px; }\n  #lean-etude .le-result-intro { font-size: 15px; margin-bottom: 30px; }\n  #lean-etude .le-question { font-size: 16px; }\n  #lean-etude .le-simple .le-result-intro { font-size: 12px; }\n  #lean-etude .le-chart-controls { flex-wrap: wrap; gap: 0; padding: 15px 14px 8px; margin-inline: -6px; }\n  #lean-etude .le-profile-buttons { width: 100%; grid-template-columns: repeat(6,minmax(0,1fr)); gap: 4px; }\n  #lean-etude .le-profile-button { min-width: 0; min-height: 44px; padding: 10px 8px; font-size: 12px; }\n  #lean-etude .le-all-button { margin-left: auto; min-height: 44px; padding: 9px 4px; font-size: 10px; }\n  #lean-etude .le-figures { margin-inline: -6px; }\n  #lean-etude .le-chart { display: block; padding: 22px; }\n  #lean-etude .le-profile-title { margin-bottom: 26px; font-size: 15px; line-height: 1.6; }\n  #lean-etude .le-profile-code { margin-bottom: 0; margin-right: 5px; width: 30px; height: 30px; font-size: 11px; }\n  #lean-etude .le-profile-description { display: inline; }\n  #lean-etude .le-chart-body { min-height: 348px; padding: 0; }\n  #lean-etude .le-axis { margin-bottom: 23px; font-size: 9px; }\n  #lean-etude .le-chart-row { gap: 12px 7px; margin-bottom: 26px; }\n  #lean-etude .le-model, #lean-etude .le-range-value { font-size: 12px; }\n  #lean-etude .le-track { height: 13px; }\n  #lean-etude .le-range::before, #lean-etude .le-range::after { top: 4.5px; }\n  #lean-etude .le-chart-caption { margin-top: 20px; padding-top: 22px; font-size: 11px; }\n  #lean-etude .le-caption-gap { font-size: 28px; margin: 13px 0; }\n  #lean-etude .le-behavior { margin-top: 40px; }\n  #lean-etude .le-behavior-title { font-size: 20px; margin-bottom: 25px; }\n  #lean-etude .le-behavior-list { grid-template-columns: 1fr; gap: 18px; }\n  #lean-etude .le-behavior-list li { grid-template-columns: 70px 1fr; grid-template-rows: 70px; align-items: center; gap: 20px; }\n  #lean-etude .le-behavior-list strong, #lean-etude .le-behavior-meter { width: 70px; height: 70px; }\n  #lean-etude .le-behavior-list strong { font-size: 23px; }\n  #lean-etude .le-behavior-list li > span:not(.le-behavior-meter) { grid-column: 2; grid-row: 1; font-size: 13px; max-width: 32ch; }\n  #lean-etude .le-insight { margin-top: 35px; padding-top: 26px; font-size: 15px; line-height: 1.75; }\n  #lean-etude .le-comparison { margin-top: 48px; }\n  #lean-etude .le-comparison-title { font-size: 26px; }\n  #lean-etude .le-comparison-stats { grid-template-columns: 1fr; border-bottom: 0; }\n  #lean-etude .le-comparison-stats li, #lean-etude .le-comparison-stats li + li { display: grid; grid-template-columns: 1fr 1fr; gap: 24px; align-items: center; padding: 25px 0; border-left: 0; border-bottom: 1px solid var(--line); }\n  #lean-etude .le-comparison-stats strong { margin: 0; font-size: 45px; }\n  #lean-etude .le-comparison-stats small { font-size: 13px; }\n  #lean-etude .le-comparison-stats span { font-size: 12px; }\n  #lean-etude .le-suivi h2 { font-size: 36px; }\n  #lean-etude .le-time-headline { margin-top: 18px; }\n  #lean-etude .le-time-intro { margin-top: 26px; margin-bottom: 45px; font-size: 14px; }\n  #lean-etude .le-time-stage { grid-template-columns: 90px minmax(0,1fr); gap: 24px; margin-bottom: 40px; }\n  #lean-etude .le-time-dial { top: calc(var(--lean-sticky-top,0px) + 100px); padding-bottom: 0; }\n  #lean-etude .le-dial-face { width: 90px; height: 90px; }\n  #lean-etude .le-dial-face::after { inset: 7px; }\n  #lean-etude .le-dial-unit { font-size: 8px; }\n  #lean-etude .le-dial-day { font-size: 44px; }\n  #lean-etude .le-dial-caption { margin-top: 17px; font-size: 9px; line-height: 1.6; }\n  #lean-etude .le-timeline li { min-height: 225px; padding: 20px 0 40px 18px; }\n  #lean-etude .le-timeline li::before { top: 25px; width: 5px; height: 5px; }\n  #lean-etude .le-timeline strong { margin-bottom: 18px; font-size: 12px; }\n  #lean-etude .le-timeline li > span { font-size: 17px; line-height: 1.6; letter-spacing: -.01em; }\n  #lean-etude .le-checklists { grid-template-columns: 1fr; gap: 40px; padding: 42px 0 56px; }\n  #lean-etude .le-checklist h3 { font-size: 27px; margin-bottom: 26px; }\n  #lean-etude .le-checklist li { font-size: 13px; }\n  #lean-etude .le-verdict { padding: 53px 22px; font-size: 28px; }\n  #lean-etude .le-verdict > strong { font-size: 11px; margin-bottom: 27px; }\n  #lean-etude .le-suivi { padding-bottom: 0; }\n  #lean-etude .le-app h2 { max-width: 18ch; font-size: 36px; }\n  #lean-etude .le-app-intro { margin-top: 28px; font-size: 15px; }\n  #lean-etude .le-app-panel { margin-top: 35px; padding: 28px 0 0; border-left: 0; border-top: 1px solid var(--line); }\n  #lean-etude .le-app-panel > strong { max-width: 24ch; font-size: 25px; }\n  #lean-etude .le-data-list { grid-template-columns: 1fr; gap: 0; }\n  #lean-etude .le-data-heading { font-size: 15px; }\n  #lean-etude .le-data-values { font-size: 12px; }\n  #lean-etude .le-disclosure summary { font-size: 16px; }\n  #lean-etude .le-disclosure p { font-size: 13px; }\n  #lean-etude .le-faq h2 { max-width: 13ch; font-size: 37px; }\n  #lean-etude .le-faq .le-disclosure:first-of-type summary { padding-top: 24px; }\n  #lean-etude .le-faq .le-disclosure:first-of-type summary::before, #lean-etude .le-faq .le-disclosure:first-of-type summary::after { top: 33px; }\n  #lean-etude .le-related { grid-template-columns: 1fr; }\n  #lean-etude .le-related a { font-size: 17px; }\n}\n@container lean-study (max-width:360px) {\n  #lean-etude .le-chart-controls { padding-inline: 6px; }\n  #lean-etude .le-suivi-inner { padding-inline: 18px; }\n  #lean-etude .le-section { padding-inline: 18px; }\n  #lean-etude .le-masthead { padding-inline: 18px; }\n  #lean-etude h1 { font-size: 36px; }\n  #lean-etude .le-title-follow { font-size: 21px; }\n  #lean-etude .le-hero-scene { height: 415px; }\n  #lean-etude .le-sculpture { height: 425px; }\n  #lean-etude .le-answer { width: 125px; padding: 9px 10px; }\n  #lean-etude .le-answer strong { font-size: 11px; }\n  #lean-etude .le-answer-0 { top: 36%; }\n  #lean-etude .le-answer-1 { top: 3%; }\n  #lean-etude .le-main-stat > strong { font-size: 56px; }\n  #lean-etude .le-main-stat > span { font-size: 12px; }\n  #lean-etude .le-hero-stats { gap: 24px 18px; }\n  #lean-etude .le-mini-stat > strong { font-size: 31px; }\n  #lean-etude .le-nav { padding-inline: 7px; }\n  #lean-etude .le-nav a { padding-inline: 7px; font-size: 9px; }\n  #lean-etude .le-method-stats { grid-template-columns: repeat(2,minmax(0,1fr)); gap: 28px; }\n  #lean-etude .le-method-stats span { font-size: 12px; }\n  #lean-etude .le-chart { padding: 18px; }\n  #lean-etude .le-model, #lean-etude .le-range-value { font-size: 11px; }\n  #lean-etude .le-caption-gap { font-size: 26px; }\n  #lean-etude .le-time-stage { grid-template-columns: 76px minmax(0,1fr); gap: 18px; }\n  #lean-etude .le-dial-face { width: 76px; height: 76px; }\n  #lean-etude .le-dial-day { font-size: 37px; }\n  #lean-etude .le-timeline li > span { font-size: 16px; }\n  #lean-etude .le-comparison-stats strong { font-size: 40px; }\n  #lean-etude .le-verdict { padding-inline: 18px; font-size: 26px; }\n}\n#lean-etude .le-suivi { padding: 0; }\n@media (prefers-reduced-motion:reduce) {\n  #lean-etude *, #lean-etude *::before, #lean-etude *::after { animation: none !important; transition: none !important; scroll-behavior: auto !important; }\n  #lean-etude .le-sculpture { transform: none; }\n}\n@media print {\n  #lean-etude { background: #fff; color: #222326; font-size: 11pt; }\n  #lean-etude .le-masthead, #lean-etude .le-hero-scene, #lean-etude .le-nav, #lean-etude .le-progress, #lean-etude .le-chart-controls, #lean-etude .le-time-dial, #lean-etude .le-hero-actions { display: none !important; }\n  #lean-etude .le-section { width: 100%; padding: 24px 0; background: #fff; color: #222326; }\n  #lean-etude .le-chart[hidden] { display: grid !important; }\n  #lean-etude .le-chart { break-inside: avoid; border: 1px solid #ccc; }\n  #lean-etude .le-ouverture, #lean-etude .le-resume, #lean-etude .le-methode, #lean-etude .le-faq, #lean-etude .le-app, #lean-etude .le-time-stage { display: block; }\n  #lean-etude .le-timeline li { min-height: 0; padding: 16px 0; }\n  #lean-etude .le-verdict { margin: 0; padding: 24px; }\n  #lean-etude .le-disclosure::details-content { content-visibility: visible; }\n}\n\n<\/style>\n<article aria-labelledby=\"le-title\" id=\"lean-etude\" lang=\"fr\"><div aria-hidden=\"true\" class=\"le-progress\"><i><\/i><\/div><section class=\"le-section le-ouverture\" id=\"le-ouverture\"><div class=\"le-hero-copy\"><p class=\"le-eyebrow\">Study \u00b7 Losing weight with AI<\/p>\n<h1 id=\"le-title\"><span class=\"le-title-line\">Losing weight<\/span> <span class=\"le-title-line\">with ChatGPT,<\/span> <span class=\"le-title-accent\">does it work?<\/span> <span class=\"le-title-follow\">We asked 5 AIs 291 times<\/span><\/h1><p class=\"le-deck\">Same person, same question, up to 1,260 kcal of difference. Here is what the AIs answer, and what they cannot do.<\/p>\n<div class=\"le-hero-actions\" data-visual-copy=\"\"><a class=\"le-explore\" href=\"#le-resume\">Explore the study<\/a><span class=\"le-scroll-cue\">\u2193<\/span><\/div><p class=\"le-byline\">Independent study by The Lean Team, the app that recomputes your expenditure every day from your measurements \u00b7 September 9, 2026 \u00b7 Open data<\/p>\n<\/div><div aria-hidden=\"true\" class=\"le-hero-scene\" data-visual-copy=\"\"><img alt=\"Buste anatomique translucide, illustration de la d\u00e9pense \u00e9nerg\u00e9tique du corps\" class=\"le-sculpture\" loading=\"eager\" decoding=\"sync\" fetchpriority=\"high\" height=\"1400\" src=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/09\/etude-chatgpt-hero-metabolism.webp\" width=\"1120\"\/><div class=\"le-scene-orbit\"><\/div><span class=\"le-scene-label\">P2 \u00b7 Simple prompt<\/span><div class=\"le-answer le-answer-0\" style=\"--order:0\"><span class=\"le-answer-name\">ChatGPT (GPT-5.5)<\/span><strong>1,600 to 2,700 kcal<\/strong><i class=\"le-answer-line\"><\/i><\/div><div class=\"le-answer le-answer-1\" style=\"--order:1\"><span class=\"le-answer-name\">ChatGPT (GPT-5.4 mini)<\/span><strong>1,600 to 2,760 kcal<\/strong><i class=\"le-answer-line\"><\/i><\/div><div class=\"le-answer le-answer-2\" style=\"--order:2\"><span class=\"le-answer-name\">Gemini 3.7 Flash<\/span><strong>1,600 to 2,600 kcal<\/strong><i class=\"le-answer-line\"><\/i><\/div><div class=\"le-answer le-answer-3\" style=\"--order:3\"><span class=\"le-answer-name\">Gemini 3.1 Pro<\/span><strong>1,600 to 2,700 kcal<\/strong><i class=\"le-answer-line\"><\/i><\/div><div class=\"le-answer le-answer-4\" style=\"--order:4\"><span class=\"le-answer-name\">Claude Opus<\/span><strong>1,500 to 2,500 kcal<\/strong><i class=\"le-answer-line\"><\/i><\/div><\/div><ul class=\"le-hero-stats\"><li class=\"le-hero-stat le-main-stat\"><strong>1 260 <span>kcal<\/span><\/strong><span>of difference for the same profile, simple prompt<\/span><\/li>\n<li class=\"le-hero-stat le-mini-stat\"><strong>0 %<\/strong><span>ask for body fat<\/span><\/li>\n<li class=\"le-hero-stat le-mini-stat\"><strong>97 %<\/strong><span>answer with a &ldquo;sedentary \/ active&rdquo; grid<\/span><\/li>\n<li class=\"le-hero-stat le-mini-stat\"><strong>14 %<\/strong><span>mention metabolic adaptation<\/span><\/li>\n<\/ul>\n<\/section>\n<nav aria-label=\"In this study\" class=\"le-nav\"><a href=\"#le-resume\">Summary<\/a><a href=\"#le-simple\">Results<\/a><a href=\"#le-suivi\">The real issue<\/a><a href=\"#le-donnees\">Data<\/a><a href=\"#le-faq\">FAQ<\/a><\/nav>\n<section aria-labelledby=\"le-resume-title\" class=\"le-section le-resume\" id=\"le-resume\"><header class=\"le-section-heading\"><span class=\"le-section-label\">Summary<\/span><h2 id=\"le-resume-title\">The study in 30 seconds<\/h2><\/header>\n<ul class=\"le-summary-list\"><li><span class=\"le-summary-num\">1<\/span><span class=\"le-summary-text\"><strong>291 answers<\/strong> from ChatGPT, Gemini and Claude to the question &ldquo;how many calories to lose weight?&rdquo;, for 6 identical profiles.<\/span><\/li>\n<li><span class=\"le-summary-num\">2<\/span><span class=\"le-summary-text\">Simple prompt (age, weight, height): the same person is told <strong>from 1,500 to 2,700 kcal<\/strong> depending on the AI and the moment.<\/span><\/li>\n<li><span class=\"le-summary-num\">3<\/span><span class=\"le-summary-text\"><strong>0 %<\/strong> of answers ask for body fat. <strong>97 %<\/strong> return a &ldquo;sedentary \/ moderate \/ active&rdquo; grid, the 1990 formula.<\/span><\/li>\n<li><span class=\"le-summary-num\">4<\/span><span class=\"le-summary-text\">Full prompt (body fat, steps, sessions): still <strong>650 kcal<\/strong> of difference, and body fat is only used in <strong>40 %<\/strong> of the calculations.<\/span><\/li>\n<li><span class=\"le-summary-num\">5<\/span><span class=\"le-summary-text\">No AI follows up the next day: steps of the day, meals, thermic effect, metabolic adaptation (<strong>14 %<\/strong> mention it). The number is frozen on day 1.<\/span><\/li>\n<\/ul>\n<div class=\"le-summary-verdict\"><strong>Verdict.<\/strong> It can work by luck, not by method: without measurement or daily updates, the answer is a population average frozen on the first day.<\/div><\/section>\n<section aria-labelledby=\"le-methode-title\" class=\"le-section le-methode\" id=\"le-methode\"><header class=\"le-section-heading\"><span class=\"le-section-label\">01 \u00b7 Method<\/span><h2 id=\"le-methode-title\">What we did<\/h2><\/header>\n<ul class=\"le-method-stats\"><li><strong>6<\/strong><span>identical profiles, from the athletic woman to the sedentary man<\/span><\/li>\n<li><strong>2<\/strong><span>prompt levels: simple, then full<\/span><\/li>\n<li><strong>5<\/strong><span>consumer AIs, fresh conversation every time<\/span><\/li>\n<li><strong>291<\/strong><span>timestamped answers, published as open data<\/span><\/li>\n<\/ul>\n<p class=\"le-method-note\">ChatGPT (GPT-5.5 and GPT-5.4 mini), Gemini (3.7 Flash and 3.1 Pro), Claude Opus. French and English, 2 to 3 repetitions, no instructions at all. The black line on the charts: the target computed component by component on measured data (BMR on lean mass, steps, sessions, TEF, 500 kcal deficit).<\/p>\n<\/section>\n<section aria-labelledby=\"le-simple-title\" class=\"le-section le-simple\" id=\"le-simple\"><header class=\"le-section-heading\"><span class=\"le-section-label\">02 \u00b7 Result 1<\/span><h2 id=\"le-simple-title\">Simple prompt: from 1,500 to 2,700 kcal for the same person<\/h2><\/header>\n<p class=\"le-result-intro\">&ldquo;I'm a 32-year-old man, 78 kg, 180 cm, I want to lose weight, how many calories a day?&rdquo; The question as millions of people ask it.<\/p>\n<div class=\"le-chart-controls\" data-controls=\"simple\" hidden=\"\"><div aria-label=\"Profiles, simple prompt\" class=\"le-profile-buttons\" role=\"group\"><button aria-controls=\"le-simple-p1\" aria-label=\"P1 \u00b7 Man, 32, 78 kg, 180 cm, 12% BF, 12,000 steps, 4 strength sessions\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"1\" type=\"button\">P1<\/button><button aria-controls=\"le-simple-p2\" aria-label=\"P2 \u00b7 Man, 32, 78 kg, 180 cm, 28% BF, 4,000 steps, no sport\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"2\" type=\"button\">P2<\/button><button aria-controls=\"le-simple-p3\" aria-label=\"P3 \u00b7 Woman, 28, 62 kg, 165 cm, 22% BF, 9,000 steps, 3 fitness sessions\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"3\" type=\"button\">P3<\/button><button aria-controls=\"le-simple-p4\" aria-label=\"P4 \u00b7 Woman, 45, 70 kg, 165 cm, 35% BF, 3,500 steps, no sport\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"4\" type=\"button\">P4<\/button><button aria-controls=\"le-simple-p5\" aria-label=\"P5 \u00b7 Man, 50, 95 kg, 178 cm, 30% BF, 6,000 steps, 2 walks\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"5\" type=\"button\">P5<\/button><button aria-controls=\"le-simple-p6\" aria-label=\"P6 \u00b7 Woman, 24, 55 kg, 170 cm, 18% BF, 14,000 steps, 5 runs\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"6\" type=\"button\">P6<\/button><\/div><button aria-pressed=\"false\" class=\"le-all-button\" data-all=\"\" type=\"button\">Show all<\/button><\/div><div class=\"le-figures\" data-charts=\"simple\"><figure aria-labelledby=\"le-simple-p1-title\" class=\"le-chart\" data-profile=\"1\" id=\"le-simple-p1\"><h3 class=\"le-profile-title\" id=\"le-simple-p1-title\"><span class=\"le-profile-code\">P1<\/span> \u00b7 <span class=\"le-profile-description\">Man, 32, 78 kg, 180 cm, 12% BF, 12,000 steps, 4 strength sessions\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"60.0\" style=\"--start:25.0%;--span:60.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 2,700<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"47.5\" style=\"--start:25.0%;--span:47.5%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 2,450<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"52.5\" style=\"--start:30.0%;--span:52.5%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,650<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.1 Pro<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"55.0\" style=\"--start:30.0%;--span:55.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,700<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"55.0\" style=\"--start:25.0%;--span:55.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 2,600<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,500 to 2,700 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">1,200 kcal spread<\/strong> (28 answers). Black line: target computed component by component on measured data (2,360 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-simple-p2-title\" class=\"le-chart\" data-profile=\"2\" id=\"le-simple-p2\"><h3 class=\"le-profile-title\" id=\"le-simple-p2-title\"><span class=\"le-profile-code\">P2<\/span> \u00b7 <span class=\"le-profile-description\">Man, 32, 78 kg, 180 cm, 28% BF, 4,000 steps, no sport<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"55.0\" style=\"--start:30.0%;--span:55.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,700<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"58.0\" style=\"--start:30.0%;--span:58.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,760<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"50.0\" style=\"--start:30.0%;--span:50.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,600<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.1 Pro<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"55.0\" style=\"--start:30.0%;--span:55.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,700<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"50.0\" style=\"--start:25.0%;--span:50.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 2,500<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,500 to 2,760 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">1,260 kcal spread<\/strong> (28 answers). Black line: target computed component by component on measured data (1,390 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-simple-p3-title\" class=\"le-chart\" data-profile=\"3\" id=\"le-simple-p3\"><h3 class=\"le-profile-title\" id=\"le-simple-p3-title\"><span class=\"le-profile-code\">P3<\/span> \u00b7 <span class=\"le-profile-description\">Woman, 28, 62 kg, 165 cm, 22% BF, 9,000 steps, 3 fitness sessions\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"40.0\" style=\"--start:10.0%;--span:40.0%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 2,000<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"6.5\" data-width=\"34.5\" style=\"--start:6.5%;--span:34.5%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,130 to 1,820<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"15.0\" data-width=\"22.5\" style=\"--start:15.0%;--span:22.5%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,300 to 1,750<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.1 Pro<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"30.0\" style=\"--start:10.0%;--span:30.0%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,800<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"35.0\" style=\"--start:10.0%;--span:35.0%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,900<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,130 to 2,000 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">870 kcal spread<\/strong> (25 answers). Black line: target computed component by component on measured data (1,499 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-simple-p4-title\" class=\"le-chart\" data-profile=\"4\" id=\"le-simple-p4\"><h3 class=\"le-profile-title\" id=\"le-simple-p4-title\"><span class=\"le-profile-code\">P4<\/span> \u00b7 <span class=\"le-profile-description\">Woman, 45, 70 kg, 165 cm, 35% BF, 3,500 steps, no sport<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"40.0\" style=\"--start:10.0%;--span:40.0%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 2,000<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"5.0\" data-width=\"37.5\" style=\"--start:5.0%;--span:37.5%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,100 to 1,850<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"12.5\" data-width=\"30.0\" style=\"--start:12.5%;--span:30.0%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,250 to 1,850<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"30.0\" style=\"--start:10.0%;--span:30.0%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,800<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,100 to 2,000 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">900 kcal spread<\/strong> (22 answers). Black line: target computed component by component on measured data (1,082 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-simple-p5-title\" class=\"le-chart\" data-profile=\"5\" id=\"le-simple-p5\"><h3 class=\"le-profile-title\" id=\"le-simple-p5-title\"><span class=\"le-profile-code\">P5<\/span> \u00b7 <span class=\"le-profile-description\">Man, 50, 95 kg, 178 cm, 30% BF, 6,000 steps, 2 walks\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"45.0\" style=\"--start:25.0%;--span:45.0%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 2,400<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"34.0\" data-width=\"43.5\" style=\"--start:34.0%;--span:43.5%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,680 to 2,550<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"32.5\" data-width=\"32.5\" style=\"--start:32.5%;--span:32.5%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,650 to 2,300<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"40.0\" style=\"--start:25.0%;--span:40.0%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 2,300<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,500 to 2,550 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">1,050 kcal spread<\/strong> (22 answers). Black line: target computed component by component on measured data (1,918 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-simple-p6-title\" class=\"le-chart\" data-profile=\"6\" id=\"le-simple-p6\"><h3 class=\"le-profile-title\" id=\"le-simple-p6-title\"><span class=\"le-profile-code\">P6<\/span> \u00b7 <span class=\"le-profile-description\">Woman, 24, 55 kg, 170 cm, 18% BF, 14,000 steps, 5 runs\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"35.0\" style=\"--start:10.0%;--span:35.0%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,900<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"32.5\" style=\"--start:10.0%;--span:32.5%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,850<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"30.0\" style=\"--start:30.0%;--span:30.0%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,200<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,200 to 2,200 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">1,000 kcal spread<\/strong> (16 answers). Black line: target computed component by component on measured data (1,735 kcal).<\/figcaption>\n<\/figure>\n<\/div><div class=\"le-behavior\"><h3 class=\"le-behavior-title\">AI behaviour, simple prompt (147 answers)<\/h3><ul class=\"le-behavior-list\"><li data-percent=\"0\" style=\"--percent:0%\"><strong>0 %<\/strong><span>ask for body fat percentage<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"1\" style=\"--percent:1%\"><strong>1 %<\/strong><span>ask for daily step count<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"97\" style=\"--percent:97%\"><strong>97 %<\/strong><span>use a declarative activity grid<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"48\" style=\"--percent:48%\"><strong>48 %<\/strong><span>cite Mifflin-St Jeor or Harris-Benedict (1990, 1919)<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"0\" style=\"--percent:0%\"><strong>0 %<\/strong><span>mention the thermic effect of food<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<\/ul>\n<\/div><div class=\"le-insight\">P1 and P2 asked <strong>exactly the same question<\/strong> (same age, weight, height) and got the same grid. Yet their real expenditure differs by nearly 1,000 kcal a day: 12% body fat and 12,000 steps on one side, 28% and 4,000 steps on the other.<\/div><\/section>\n<section aria-labelledby=\"le-complet-title\" class=\"le-section le-complet\" id=\"le-complet\"><header class=\"le-section-heading\"><span class=\"le-section-label\">03 \u00b7 Result 2<\/span><h2 id=\"le-complet-title\">Full prompt: still a 650 kcal gap<\/h2><\/header>\n<p class=\"le-result-intro\">We give everything: body fat, steps per day, sessions, and &ldquo;give me a precise number&rdquo;. All comply. A precise number is not a correct number.<\/p>\n<div class=\"le-chart-controls\" data-controls=\"complet\" hidden=\"\"><div aria-label=\"Profiles, full prompt\" class=\"le-profile-buttons\" role=\"group\"><button aria-controls=\"le-complet-p1\" aria-label=\"P1 \u00b7 Man, 32, 78 kg, 180 cm, 12% BF, 12,000 steps, 4 strength sessions\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"1\" type=\"button\">P1<\/button><button aria-controls=\"le-complet-p2\" aria-label=\"P2 \u00b7 Man, 32, 78 kg, 180 cm, 28% BF, 4,000 steps, no sport\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"2\" type=\"button\">P2<\/button><button aria-controls=\"le-complet-p3\" aria-label=\"P3 \u00b7 Woman, 28, 62 kg, 165 cm, 22% BF, 9,000 steps, 3 fitness sessions\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"3\" type=\"button\">P3<\/button><button aria-controls=\"le-complet-p4\" aria-label=\"P4 \u00b7 Woman, 45, 70 kg, 165 cm, 35% BF, 3,500 steps, no sport\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"4\" type=\"button\">P4<\/button><button aria-controls=\"le-complet-p5\" aria-label=\"P5 \u00b7 Man, 50, 95 kg, 178 cm, 30% BF, 6,000 steps, 2 walks\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"5\" type=\"button\">P5<\/button><button aria-controls=\"le-complet-p6\" aria-label=\"P6 \u00b7 Woman, 24, 55 kg, 170 cm, 18% BF, 14,000 steps, 5 runs\/wk\" aria-pressed=\"false\" class=\"le-profile-button\" data-profile=\"6\" type=\"button\">P6<\/button><\/div><button aria-pressed=\"false\" class=\"le-all-button\" data-all=\"\" type=\"button\">Show all<\/button><\/div><div class=\"le-figures\" data-charts=\"complet\"><figure aria-labelledby=\"le-complet-p1-title\" class=\"le-chart\" data-profile=\"1\" id=\"le-complet-p1\"><h3 class=\"le-profile-title\" id=\"le-complet-p1-title\"><span class=\"le-profile-code\">P1<\/span> \u00b7 <span class=\"le-profile-description\">Man, 32, 78 kg, 180 cm, 12% BF, 12,000 steps, 4 strength sessions\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"67.5\" data-width=\"20.0\" style=\"--start:67.5%;--span:20.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">2,350 to 2,750<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"65.0\" data-width=\"10.0\" style=\"--start:65.0%;--span:10.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">2,300 to 2,500<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"55.0\" data-width=\"15.0\" style=\"--start:55.0%;--span:15.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">2,100 to 2,400<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.1 Pro<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"65.0\" data-width=\"20.0\" style=\"--start:65.0%;--span:20.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">2,300 to 2,700<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"62.5\" data-width=\"10.0\" style=\"--start:62.5%;--span:10.0%\"><\/div><span class=\"le-target\" data-target=\"68.0\" style=\"--target:68.0%\"><\/span><\/div><span class=\"le-range-value\">2,250 to 2,450<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">2,100 to 2,750 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">650 kcal spread<\/strong> (28 answers). Black line: target computed component by component on measured data (2,360 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-complet-p2-title\" class=\"le-chart\" data-profile=\"2\" id=\"le-complet-p2\"><h3 class=\"le-profile-title\" id=\"le-complet-p2-title\"><span class=\"le-profile-code\">P2<\/span> \u00b7 <span class=\"le-profile-description\">Man, 32, 78 kg, 180 cm, 28% BF, 4,000 steps, no sport<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"10.0\" style=\"--start:30.0%;--span:10.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 1,800<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"37.5\" data-width=\"12.5\" style=\"--start:37.5%;--span:12.5%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,750 to 2,000<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"27.5\" data-width=\"10.0\" style=\"--start:27.5%;--span:10.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,550 to 1,750<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.1 Pro<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"20.0\" data-width=\"20.0\" style=\"--start:20.0%;--span:20.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,400 to 1,800<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"32.5\" data-width=\"15.0\" style=\"--start:32.5%;--span:15.0%\"><\/div><span class=\"le-target\" data-target=\"19.5\" style=\"--target:19.5%\"><\/span><\/div><span class=\"le-range-value\">1,650 to 1,950<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,400 to 2,000 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">600 kcal spread<\/strong> (28 answers). Black line: target computed component by component on measured data (1,390 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-complet-p3-title\" class=\"le-chart\" data-profile=\"3\" id=\"le-complet-p3\"><h3 class=\"le-profile-title\" id=\"le-complet-p3-title\"><span class=\"le-profile-code\">P3<\/span> \u00b7 <span class=\"le-profile-description\">Woman, 28, 62 kg, 165 cm, 22% BF, 9,000 steps, 3 fitness sessions\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"12.5\" style=\"--start:30.0%;--span:12.5%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 1,850<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"35.0\" data-width=\"7.5\" style=\"--start:35.0%;--span:7.5%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,700 to 1,850<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"25.0\" data-width=\"12.5\" style=\"--start:25.0%;--span:12.5%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,500 to 1,750<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"22.5\" data-width=\"22.5\" style=\"--start:22.5%;--span:22.5%\"><\/div><span class=\"le-target\" data-target=\"24.9\" style=\"--target:24.9%\"><\/span><\/div><span class=\"le-range-value\">1,450 to 1,900<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,450 to 1,900 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">450 kcal spread<\/strong> (22 answers). Black line: target computed component by component on measured data (1,499 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-complet-p4-title\" class=\"le-chart\" data-profile=\"4\" id=\"le-complet-p4\"><h3 class=\"le-profile-title\" id=\"le-complet-p4-title\"><span class=\"le-profile-code\">P4<\/span> \u00b7 <span class=\"le-profile-description\">Woman, 45, 70 kg, 165 cm, 35% BF, 3,500 steps, no sport<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"12.5\" style=\"--start:10.0%;--span:12.5%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,450<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"12.5\" data-width=\"12.5\" style=\"--start:12.5%;--span:12.5%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,250 to 1,500<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"15.0\" data-width=\"5.0\" style=\"--start:15.0%;--span:5.0%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,300 to 1,400<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"10.0\" data-width=\"20.0\" style=\"--start:10.0%;--span:20.0%\"><\/div><span class=\"le-target\" data-target=\"4.1\" style=\"--target:4.1%\"><\/span><\/div><span class=\"le-range-value\">1,200 to 1,600<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,200 to 1,600 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">400 kcal spread<\/strong> (22 answers). Black line: target computed component by component on measured data (1,082 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-complet-p5-title\" class=\"le-chart\" data-profile=\"5\" id=\"le-complet-p5\"><h3 class=\"le-profile-title\" id=\"le-complet-p5-title\"><span class=\"le-profile-code\">P5<\/span> \u00b7 <span class=\"le-profile-description\">Man, 50, 95 kg, 178 cm, 30% BF, 6,000 steps, 2 walks\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"42.5\" data-width=\"17.5\" style=\"--start:42.5%;--span:17.5%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,850 to 2,200<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"45.0\" data-width=\"7.5\" style=\"--start:45.0%;--span:7.5%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,900 to 2,050<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"42.5\" data-width=\"6.5\" style=\"--start:42.5%;--span:6.5%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,850 to 1,980<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"30.0\" data-width=\"25.0\" style=\"--start:30.0%;--span:25.0%\"><\/div><span class=\"le-target\" data-target=\"45.9\" style=\"--target:45.9%\"><\/span><\/div><span class=\"le-range-value\">1,600 to 2,100<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,600 to 2,200 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">600 kcal spread<\/strong> (22 answers). Black line: target computed component by component on measured data (1,918 kcal).<\/figcaption>\n<\/figure>\n<figure aria-labelledby=\"le-complet-p6-title\" class=\"le-chart\" data-profile=\"6\" id=\"le-complet-p6\"><h3 class=\"le-profile-title\" id=\"le-complet-p6-title\"><span class=\"le-profile-code\">P6<\/span> \u00b7 <span class=\"le-profile-description\">Woman, 24, 55 kg, 170 cm, 18% BF, 14,000 steps, 5 runs\/wk<\/span><\/h3><div class=\"le-chart-body\"><div class=\"le-axis\"><span>1,000 kcal<\/span><span>3,000 kcal<\/span><div aria-hidden=\"true\" class=\"le-axis-marks\"><i><\/i><i><\/i><i><\/i><i><\/i><i><\/i><\/div><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.5)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"45.0\" data-width=\"20.0\" style=\"--start:45.0%;--span:20.0%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">1,900 to 2,300<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">ChatGPT (GPT-5.4 mini)<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"42.5\" data-width=\"17.5\" style=\"--start:42.5%;--span:17.5%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">1,850 to 2,200<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Gemini 3.7 Flash<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"65.0\" data-width=\"10.0\" style=\"--start:65.0%;--span:10.0%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">2,300 to 2,500<\/span><\/div><div class=\"le-chart-row\"><span class=\"le-model\">Claude Opus<\/span><div aria-hidden=\"true\" class=\"le-track\"><div class=\"le-range\" data-start=\"45.0\" data-width=\"25.0\" style=\"--start:45.0%;--span:25.0%\"><\/div><span class=\"le-target\" data-target=\"36.8\" style=\"--target:36.8%\"><\/span><\/div><span class=\"le-range-value\">1,900 to 2,400<\/span><\/div><\/div><figcaption class=\"le-chart-caption\">All AIs combined: <strong class=\"le-caption-range\">1,850 to 2,500 kcal<\/strong>, i.e. <strong class=\"le-caption-gap\">650 kcal spread<\/strong> (19 answers). Black line: target computed component by component on measured data (1,735 kcal).<\/figcaption>\n<\/figure>\n<\/div><div class=\"le-behavior\"><h3 class=\"le-behavior-title\">AI behaviour, full prompt (144 answers)<\/h3><ul class=\"le-behavior-list\"><li data-percent=\"40\" style=\"--percent:40%\"><strong>40 %<\/strong><span>actually use the provided body fat<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"42\" style=\"--percent:42%\"><strong>42 %<\/strong><span>convert the provided steps into calories<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"14\" style=\"--percent:14%\"><strong>14 %<\/strong><span>mention metabolic adaptation<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"3\" style=\"--percent:3%\"><strong>3 %<\/strong><span>mention the thermic effect of food<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<li data-percent=\"97\" style=\"--percent:97%\"><strong>97 %<\/strong><span>give a single number, as requested<\/span><span aria-hidden=\"true\" class=\"le-behavior-meter\"><i><\/i><\/span><\/li>\n<\/ul>\n<\/div><div class=\"le-comparison\" data-comparison=\"\"><h3 class=\"le-comparison-title\">The P1 vs P2 test: same weight, different bodies<\/h3><ul class=\"le-comparison-stats\"><li><strong>699 <small>kcal<\/small><\/strong><span>of difference between P1 and P2 according to the AIs (average)<\/span><\/li>\n<li><strong>970 <small>kcal<\/small><\/strong><span>of difference when computing component by component<\/span><\/li>\n<li><strong>+324 <small>kcal<\/small><\/strong><span>too much for the sedentary profile at 28% body fat<\/span><\/li>\n<\/ul>\n<\/div><div class=\"le-insight\">The AIs see the difference between the two men, but underestimate it: they adjust a population formula instead of measuring. The sedentary profile, the one who &ldquo;does everything right&rdquo;, gets a target that is too high and loses nothing.<\/div><\/section>\n<section aria-labelledby=\"le-suivi-title\" class=\"le-section le-suivi\" id=\"le-suivi\"><div class=\"le-suivi-inner\"><header class=\"le-section-heading\"><span class=\"le-section-label\">04 \u00b7 The real issue<\/span><h2 id=\"le-suivi-title\">The problem is not day 1, it is the <span class=\"le-time-headline\">89 days that follow<\/span><\/h2><\/header>\n<p class=\"le-time-intro\">78% of detailed answers advise to &ldquo;weigh yourself and adjust&rdquo;. None does it for you.<\/p>\n<div class=\"le-time-stage\"><div aria-hidden=\"true\" class=\"le-time-dial\" data-visual-copy=\"\"><div class=\"le-dial-face\"><span class=\"le-dial-unit\" data-current-unit=\"\">Day<\/span><span class=\"le-dial-day\" data-current-day=\"\">1<\/span><\/div><span class=\"le-dial-caption\">The number is frozen on day 1.<\/span><\/div><ul class=\"le-timeline\"><li><strong><span>Day<\/span> <b>1<\/b><\/strong><span>The chatbot receives a weight and answers once.<\/span><\/li>\n<li><strong><span>Day<\/span> <b>2<\/b><\/strong><span>You walked 4,000 steps instead of 11,000. It does not know.<\/span><\/li>\n<li><strong><span>Week<\/span> <b>3<\/b><\/strong><span>You did 6 sessions and changed your macros. It does not know.<\/span><\/li>\n<li><strong><span>Day<\/span> <b>90<\/b><\/strong><span>You lost 4 kg, your metabolism dropped then adapted. It still does not know.<\/span><\/li>\n<\/ul>\n<\/div><div class=\"le-checklists\"><div class=\"le-checklist le-checklist-no\"><h3>What the chatbot does not know<\/h3><ul><li><span>\u2715<\/span><span>Your steps of the day<\/span><\/li>\n<li><span>\u2715<\/span><span>Your meals of the day and their thermic effect<\/span><\/li>\n<li><span>\u2715<\/span><span>Your real body fat, and how it changes<\/span><\/li>\n<li><span>\u2715<\/span><span>How your metabolism adapts over the weeks<\/span><\/li>\n<li><span>\u2715<\/span><span>No curve, no chart of your expenditure<\/span><\/li>\n<\/ul>\n<\/div><div class=\"le-checklist le-checklist-yes\"><h3>What real tracking requires<\/h3><ul><li><span>\u2713<\/span><span>Expenditure recomputed every day from measurements<\/span><\/li>\n<li><span>\u2713<\/span><span>A target that moves with steps and sessions<\/span><\/li>\n<li><span>\u2713<\/span><span>Body fat re-measured regularly<\/span><\/li>\n<li><span>\u2713<\/span><span>Automatic correction of metabolic adaptation<\/span><\/li>\n<li><span>\u2713<\/span><span>A readable curve of your TDEE and its evolution<\/span><\/li>\n<\/ul>\n<\/div><\/div><\/div><div class=\"le-verdict\"><strong>Verdict: does losing weight with ChatGPT work?<\/strong> By luck, sometimes. By method, no: without measurement or daily updates, the answer remains a population estimate frozen on the first day. The test can be redone in five minutes with your own profile.<\/div><\/section>\n<section aria-labelledby=\"le-app-title\" class=\"le-section le-app\" id=\"le-app\"><header class=\"le-section-heading\"><span class=\"le-section-label\">05 \u00b7 And Lean?<\/span><h2 id=\"le-app-title\">The AI that measures versus the AI that guesses<\/h2><\/header>\n<p class=\"le-app-intro\">Lean also uses AI, but to measure: BodyScan for body fat, photo scan for the plate, pedometer for steps. Then it recomputes the five components of your expenditure every day and plots the curve.<\/p>\n<div class=\"le-app-panel\"><div class=\"le-feature-list\"><span class=\"le-feature le-feature-0\">BMR on measured lean mass<\/span><span class=\"le-feature le-feature-1\">NEAT on real steps<\/span><span class=\"le-feature le-feature-2\">EAT per session<\/span><span class=\"le-feature le-feature-3\">TEF from macros<\/span><span class=\"le-feature le-feature-4\">Metabolic adaptation<\/span><\/div><strong>Lean measures, recomputes and tracks, every day<\/strong><span>First app to compute metabolic adaptation. Free download, 7-day free trial on the annual subscription. 4.7\/5 on the App Store, more than 10,000 users.<\/span><p class=\"le-store-links\"><a href=\"https:\/\/apps.apple.com\/fr\/app\/lean-calorie-ai-podometre\/id6738668646?utm_source=seo&amp;utm_medium=etude&amp;utm_campaign=chatgpt-calories\" rel=\"noopener\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-appstore-official.webp\" alt=\"T\u00e9l\u00e9charger Lean sur l\\'App Store\" width=\"150\" height=\"50\" style=\"height:50px;width:auto;display:block;\"><\/a><a href=\"https:\/\/play.google.com\/store\/apps\/details?id=com.lean.testsqflite&amp;utm_source=seo&amp;utm_medium=etude&amp;utm_campaign=chatgpt-calories\" rel=\"noopener\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-googleplay-official.webp\" alt=\"T\u00e9l\u00e9charger Lean sur Google Play\" width=\"168\" height=\"50\" style=\"height:50px;width:auto;display:block;\"><\/a><\/p>\n<\/div><\/section>\n<section aria-labelledby=\"le-donnees-title\" class=\"le-section le-donnees\" id=\"le-donnees\"><header class=\"le-section-heading\"><span class=\"le-section-label\">Data<\/span><h2 id=\"le-donnees-title\">The data, profile by profile<\/h2><\/header>\n<ul class=\"le-data-list\"><li class=\"le-data-record\" id=\"le-data-p1\"><h3 class=\"le-data-heading\"><strong>P1<\/strong> \u00b7 Man, 32, 78 kg, 180 cm, 12% BF, 12,000 steps, 4 strength sessions\/wk<\/h3><div class=\"le-data-values\">Simple prompt: <strong>1,500 to 2,700 kcal<\/strong> (28 answers) \u00b7 Full prompt: <strong>2,100 to 2,750 kcal<\/strong> (28 answers) \u00b7 Component-by-component target: <strong>2,360 kcal<\/strong> (BMR 1,853, NEAT 504, EAT 267, TEF 236, TDEE 2,860)<\/div><\/li>\n<li class=\"le-data-record\" id=\"le-data-p2\"><h3 class=\"le-data-heading\"><strong>P2<\/strong> \u00b7 Man, 32, 78 kg, 180 cm, 28% BF, 4,000 steps, no sport<\/h3><div class=\"le-data-values\">Simple prompt: <strong>1,500 to 2,760 kcal<\/strong> (28 answers) \u00b7 Full prompt: <strong>1,400 to 2,000 kcal<\/strong> (28 answers) \u00b7 Component-by-component target: <strong>1,390 kcal<\/strong> (BMR 1,583, NEAT 168, EAT 0, TEF 139, TDEE 1,890)<\/div><\/li>\n<li class=\"le-data-record\" id=\"le-data-p3\"><h3 class=\"le-data-heading\"><strong>P3<\/strong> \u00b7 Woman, 28, 62 kg, 165 cm, 22% BF, 9,000 steps, 3 fitness sessions\/wk<\/h3><div class=\"le-data-values\">Simple prompt: <strong>1,130 to 2,000 kcal<\/strong> (25 answers) \u00b7 Full prompt: <strong>1,450 to 1,900 kcal<\/strong> (22 answers) \u00b7 Component-by-component target: <strong>1,499 kcal<\/strong> (BMR 1,415, NEAT 275, EAT 159, TEF 150, TDEE 1,999)<\/div><\/li>\n<li class=\"le-data-record\" id=\"le-data-p4\"><h3 class=\"le-data-heading\"><strong>P4<\/strong> \u00b7 Woman, 45, 70 kg, 165 cm, 35% BF, 3,500 steps, no sport<\/h3><div class=\"le-data-values\">Simple prompt: <strong>1,100 to 2,000 kcal<\/strong> (22 answers) \u00b7 Full prompt: <strong>1,200 to 1,600 kcal<\/strong> (22 answers) \u00b7 Component-by-component target: <strong>1,082 kcal<\/strong> (BMR 1,353, NEAT 121, EAT 0, TEF 108, TDEE 1,582)<\/div><\/li>\n<li class=\"le-data-record\" id=\"le-data-p5\"><h3 class=\"le-data-heading\"><strong>P5<\/strong> \u00b7 Man, 50, 95 kg, 178 cm, 30% BF, 6,000 steps, 2 walks\/wk<\/h3><div class=\"le-data-values\">Simple prompt: <strong>1,500 to 2,550 kcal<\/strong> (22 answers) \u00b7 Full prompt: <strong>1,600 to 2,200 kcal<\/strong> (22 answers) \u00b7 Component-by-component target: <strong>1,918 kcal<\/strong> (BMR 1,806, NEAT 303, EAT 117, TEF 192, TDEE 2,418)<\/div><\/li>\n<li class=\"le-data-record\" id=\"le-data-p6\"><h3 class=\"le-data-heading\"><strong>P6<\/strong> \u00b7 Woman, 24, 55 kg, 170 cm, 18% BF, 14,000 steps, 5 runs\/wk<\/h3><div class=\"le-data-values\">Simple prompt: <strong>1,200 to 2,200 kcal<\/strong> (16 answers) \u00b7 Full prompt: <strong>1,850 to 2,500 kcal<\/strong> (19 answers) \u00b7 Component-by-component target: <strong>1,735 kcal<\/strong> (BMR 1,344, NEAT 391, EAT 326, TEF 174, TDEE 2,235)<\/div><\/li>\n<\/ul>\n<p class=\"le-data-note\">Variability of the same AI on the same repeated prompt (full prompt, average gap between repetitions): Claude Opus 73 kcal, Gemini 3.1 Pro 175 kcal, Gemini 3.7 Flash 57 kcal, ChatGPT (GPT-5.4 mini) 98 kcal, ChatGPT (GPT-5.5) 96 kcal.<\/p>\n<p class=\"le-data-note\">Full dataset (prompts, 291 raw answers, extraction, scripts): <a href=\"https:\/\/doi.org\/10.5281\/zenodo.22662006\" rel=\"noopener\">Zenodo, DOI 10.5281\/zenodo.22662006<\/a>, CC BY 4.0 licence.<\/p>\n<details class=\"le-disclosure\"><summary>Detailed method and limitations<\/summary>\n<p>Models queried by API with default parameters and no system instruction: GPT-5.5 and GPT-5.4 mini (OpenAI), Gemini 3.7 Flash and Gemini 3.1 Pro (Google). Claude Opus (Anthropic) queried through the official command-line interface with a neutral system prompt and no context. Six profiles, two prompt levels, French and English, 2 to 3 repetitions per combination, 291 answers collected on September 8, 2026 (Gemini 3.1 Pro: 27 answers, API rate limited; other models: 48 to 72).<\/p>\n<p>Reading the answers: an automatic extractor (GPT-5.4 mini, temperature 0, strict JSON schema) identifies the recommended intake target (min and max when a range or a grid is given) and the behaviours (question asked before answering, activity grid, actual use of the provided body fat and steps, mention of metabolic adaptation and of the thermic effect). The extraction is published with the raw answers, line by line.<\/p>\n<p>Limitations: fictional profiles; public reference formulas (Katch-McArdle on lean mass, Ainsworth 2011 compendium, TEF at 10%) that serve as a benchmark, not ground truth; models evolve and answers change from one week to the next, which is part of the finding. Lean publishes this study and sells a tracking app: method and data are open so that anyone can redo it.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>Scientific sources<\/summary>\n<ul><li>Mifflin MD et al. A new predictive equation for resting energy expenditure in healthy individuals. <em>Am J Clin Nutr<\/em>. 1990;51(2):241-7. <a href=\"https:\/\/pubmed.ncbi.nlm.nih.gov\/2305711\/\" rel=\"noopener\">PubMed 2305711<\/a><\/li>\n<li>Ainsworth BE et al. 2011 Compendium of Physical Activities. <em>Med Sci Sports Exerc<\/em>. 2011;43(8):1575-81. <a href=\"https:\/\/pubmed.ncbi.nlm.nih.gov\/21681120\/\" rel=\"noopener\">PubMed 21681120<\/a><\/li>\n<li>Levine JA et al. Interindividual variation in posture allocation. <em>Science<\/em>. 2005;307(5709):584-6. <a href=\"https:\/\/pubmed.ncbi.nlm.nih.gov\/15681386\/\" rel=\"noopener\">PubMed 15681386<\/a><\/li>\n<li>Trexler ET et al. Metabolic adaptation to weight loss: implications for the athlete. <em>J Int Soc Sports Nutr<\/em>. 2014;11:7. <a href=\"https:\/\/pubmed.ncbi.nlm.nih.gov\/24571926\/\" rel=\"noopener\">PubMed 24571926<\/a><\/li>\n<li>Westerterp KR. Diet induced thermogenesis. <em>Nutr Metab<\/em>. 2004;1:5. <a href=\"https:\/\/pubmed.ncbi.nlm.nih.gov\/15507147\/\" rel=\"noopener\">PubMed 15507147<\/a><\/li>\n<\/ul>\n<\/details>\n<p class=\"le-data-note\">Study published by The Lean Team. Press contact: franckbarriere@lean-app.com. Not medical advice.<\/p>\n<\/section>\n<section aria-labelledby=\"le-faq-title\" class=\"le-section le-faq\" id=\"le-faq\"><header class=\"le-section-heading\"><h2 id=\"le-faq-title\">Frequently asked questions<\/h2><\/header>\n<details class=\"le-disclosure\"><summary>Can ChatGPT calculate how many calories I should eat?<\/summary>\n<p>It gives a number, not a measurement. Out of 291 answers, no AI asked for body fat (a single one asked for the step count), and 97% of answers to the simple prompt rely on a declarative activity grid (sedentary, moderate, active). The result is a statistical average that ignores your body composition and your real activity, with up to 1,260 kcal of difference for the same person.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>Why do AIs give such different numbers for the same profile?<\/summary>\n<p>Because they have no measurement. They apply a population formula (most often Mifflin-St Jeor 1990) then an activity multiplier picked by guesswork. Two 78 kg men, one at 12% body fat and 12,000 steps, the other at 28% and 4,000 steps, get the same answer to the simple prompt, while their real expenditure differs by nearly 1,000 kcal.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>Does giving ChatGPT more details fix the problem?<\/summary>\n<p>Only partly. With body fat, steps and sessions in the prompt, the gap between AIs remains 650 kcal on the same profile, body fat is actually used in the calculation in only 40% of answers and steps in 42%. Above all, the answer is frozen: it does not know how much you walked or ate today, nor how your metabolism adapts over the weeks.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>What is metabolic adaptation and why does it change the answer?<\/summary>\n<p>In a prolonged deficit, basal metabolic rate drops beyond what the weight loss explains. A number that is right today becomes too high after a few weeks. Only 14% of detailed answers mentioned it, and none computes it. It is the most common cause of plateaus.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>How do you correctly calculate your calories to lose weight?<\/summary>\n<p>By rebuilding expenditure component by component, TDEE = BMR + NEAT + EAT + TEF, with basal metabolic rate computed on real lean mass, NEAT on measured steps, EAT on sessions actually done, TEF on macros eaten, then tracking the evolution day after day and correcting when the body adapts. It is tracking, not a one-sentence answer.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>Is AI useless for nutrition?<\/summary>\n<p>No. AI is valuable when it measures: estimating body fat from a photo, recognising a meal, reading a barcode. It is weak when it guesses from three numbers. The difference is between an AI plugged into continuously updated measurements and a chatbot answering a stranger from memory.<\/p>\n<\/details>\n<details class=\"le-disclosure\"><summary>Can I reproduce the study?<\/summary>\n<p>Yes. The 12 prompts, the timestamped raw answers, the structured extraction and the scripts are published as open data (link at the bottom of the page). Each call is a fresh conversation, with no system instruction, with default parameters.<\/p>\n<\/details>\n<\/section>\n<section aria-labelledby=\"le-suite-title\" class=\"le-section le-suite\" id=\"le-suite\"><header class=\"le-section-heading\"><h2 id=\"le-suite-title\">Read also<\/h2><\/header>\n<ul class=\"le-related\"><li><a href=\"https:\/\/lean-app.com\/en\/calculateur-tdee\/\">TDEE Calculator: the canonical formula BMR + NEAT + EAT + TEF <span>Bodyfat-aware calculator with breakdown of the 4 metabolic components.<\/span><\/a><\/li>\n<li><a href=\"https:\/\/lean-app.com\/en\/calculer-vraie-depense-calorique\/\">How to calculate your real daily calorie burn: the complete guide <span>BMR on actual body fat, NEAT from steps, EAT per session, TEF from macros, adaptation.<\/span><\/a><\/li>\n<li><a href=\"https:\/\/lean-app.com\/en\/calculateur-adaptation-metabolique\/\">Metabolic adaptation calculator: has your metabolism slowed down? <span>Coefficient from 100 to 0 %, adapted BMR in kcal\/day, and when to take a diet break.<\/span><\/a><\/li>\n<li><a href=\"https:\/\/lean-app.com\/en\/10000-pas-calories\/\">10,000 steps a day: how many calories it really burns <span>Between 300 and 550 kcal depending on your weight. Table by profile, public formula, calculator.<\/span><\/a><\/li>\n<li><a href=\"https:\/\/lean-app.com\/en\/etude-base-donnees-calories\/\">Study: 857,655 products analysed, 1 scan in 3 unreliable <span>The quality of calorie databases, measured.<\/span><\/a><\/li>\n<\/ul>\n<\/section>\n<script type=\"application\/ld+json\">{\"@context\": \"https:\/\/schema.org\", \"@type\": \"FAQPage\", \"mainEntity\": [{\"@type\": \"Question\", \"name\": \"ChatGPT peut-il calculer combien de calories je dois manger ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"Il donne un chiffre, mais pas une mesure. Sur 291 r\u00e9ponses, aucune IA n'a demand\u00e9 la masse grasse (une seule a demand\u00e9 le nombre de pas), et 97 % des r\u00e9ponses au prompt simple reposent sur une grille d'activit\u00e9 d\u00e9clarative (s\u00e9dentaire, mod\u00e9r\u00e9, actif). Le r\u00e9sultat est une moyenne statistique qui ignore ta composition corporelle et ton activit\u00e9 r\u00e9elle, avec jusqu'\u00e0 1 260 kcal d'\u00e9cart pour la m\u00eame personne.\"}}, {\"@type\": \"Question\", \"name\": \"Pourquoi les IA donnent-elles des chiffres si diff\u00e9rents pour le m\u00eame profil ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"Parce qu'elles n'ont pas de mesure. Elles appliquent une formule de population (Mifflin-St Jeor 1990 le plus souvent) puis un multiplicateur d'activit\u00e9 choisi au jug\u00e9. Deux hommes de 78 kg, l'un \u00e0 12 % de masse grasse et 12 000 pas, l'autre \u00e0 28 % et 4 000 pas, re\u00e7oivent la m\u00eame r\u00e9ponse au prompt simple, alors que leur d\u00e9pense r\u00e9elle diff\u00e8re de pr\u00e8s de 1 000 kcal.\"}}, {\"@type\": \"Question\", \"name\": \"Est-ce que donner plus de d\u00e9tails \u00e0 ChatGPT corrige le probl\u00e8me ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"En partie seulement. Avec la masse grasse, les pas et les s\u00e9ances dans le prompt, l'\u00e9cart entre IA reste de 650 kcal sur un m\u00eame profil, la masse grasse n'est r\u00e9ellement utilis\u00e9e dans le calcul que dans 40 % des r\u00e9ponses et les pas dans 42 %. Et surtout, la r\u00e9ponse est fig\u00e9e : elle ne sait pas ce que tu as march\u00e9 ni mang\u00e9 aujourd'hui, ni comment ton m\u00e9tabolisme s'adapte au fil des semaines.\"}}, {\"@type\": \"Question\", \"name\": \"Qu'est-ce que l'adaptation m\u00e9tabolique et pourquoi \u00e7a change la r\u00e9ponse ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"En d\u00e9ficit prolong\u00e9, le m\u00e9tabolisme de base baisse au-del\u00e0 de ce que la perte de poids explique. Un chiffre juste aujourd'hui devient trop haut au bout de quelques semaines. Seules 14 % des r\u00e9ponses d\u00e9taill\u00e9es l'ont mentionn\u00e9, et aucune ne la calcule. C'est la cause la plus fr\u00e9quente des plateaux.\"}}, {\"@type\": \"Question\", \"name\": \"Comment calculer correctement ses calories pour maigrir ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"En reconstruisant la d\u00e9pense brique par brique, TDEE = BMR + NEAT + EAT + TEF, avec le m\u00e9tabolisme de base calcul\u00e9 sur la masse maigre r\u00e9elle, la NEAT sur les pas mesur\u00e9s, l'EAT sur les s\u00e9ances r\u00e9ellement faites, le TEF sur les macros ing\u00e9r\u00e9s, puis en suivant l'\u00e9volution jour apr\u00e8s jour et en corrigeant quand le corps s'adapte. C'est un suivi, pas une r\u00e9ponse en une phrase.\"}}, {\"@type\": \"Question\", \"name\": \"L'IA est-elle inutile pour la nutrition ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"Non. L'IA est pr\u00e9cieuse quand elle mesure : estimer une masse grasse sur une photo, reconna\u00eetre un plat, lire un code-barres. Elle est faible quand elle devine \u00e0 partir de trois chiffres. La diff\u00e9rence est entre une IA branch\u00e9e sur des mesures mises \u00e0 jour en continu et un chatbot qui r\u00e9pond de m\u00e9moire \u00e0 un inconnu.\"}}, {\"@type\": \"Question\", \"name\": \"Puis-je reproduire l'\u00e9tude ?\", \"acceptedAnswer\": {\"@type\": \"Answer\", \"text\": \"Oui. Les 12 prompts, les r\u00e9ponses brutes horodat\u00e9es, l'extraction structur\u00e9e et les scripts sont publi\u00e9s en donn\u00e9es ouvertes (lien en bas de page). Chaque appel est une conversation neuve, sans instruction syst\u00e8me, avec les param\u00e8tres par d\u00e9faut.\"}}]}<\/script><script type=\"application\/ld+json\">{\"@context\": \"https:\/\/schema.org\", \"@type\": \"Dataset\", \"name\": \"Asking consumer AI assistants how many calories to eat to lose weight: 5 models, 6 profiles, 291 answers\", \"description\": \"On a demand\u00e9 291 fois \u00e0 ChatGPT, Gemini et Claude combien manger pour maigrir, pour 6 profils identiques. Aucune IA ne demande la masse grasse ni les pas, 97 % r\u00e9pondent par une grille d'activit\u00e9, et le m\u00eame profil re\u00e7oit jusqu'\u00e0 1260 kcal d'\u00e9cart. M\u00e9thode, donn\u00e9es brutes et ce qu'il faudrait faire \u00e0 la place.\", \"url\": \"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/\", \"identifier\": \"https:\/\/doi.org\/10.5281\/zenodo.22662006\", \"license\": \"https:\/\/creativecommons.org\/licenses\/by\/4.0\/\", \"creator\": {\"@type\": \"Organization\", \"name\": \"Lean\", \"url\": \"https:\/\/lean-app.com\/\"}, \"datePublished\": \"2026-09-08\", \"keywords\": [\"ChatGPT\", \"Gemini\", \"Claude\", \"calories\", \"TDEE\", \"weight loss\", \"metabolic adaptation\"], \"variableMeasured\": [\"recommended daily calorie intake (kcal)\", \"asks for body fat\", \"uses activity factor\", \"mentions metabolic adaptation\", \"mentions thermic effect of food\"], \"distribution\": {\"@type\": \"DataDownload\", \"encodingFormat\": \"application\/zip\", \"contentUrl\": \"https:\/\/zenodo.org\/records\/22662006\"}}<\/script><\/article>\n<script data-wpmeteor-nooptimize=\"true\">\n(function () {\n  'use strict';\n  var root = document.getElementById('lean-etude');\n  if (!root || root.dataset.enhanced) return;\n  root.dataset.enhanced = 'true';\n  var reduce = window.matchMedia('(prefers-reduced-motion: reduce)');\n  var animate = function (element, frames, options) {\n    if (!(reduce.matches || !element.animate)) element.animate(frames, options);\n  };\n  root.querySelectorAll('[data-controls]').forEach(function (controls) {\n    var group = root.querySelector('[data-charts=\"' + controls.dataset.controls + '\"]');\n    if (!group) return;\n    var figures = Array.from(group.querySelectorAll('.le-chart'));\n    var buttons = Array.from(controls.querySelectorAll('[data-profile]'));\n    var allButton = controls.querySelector('[data-all]');\n    var select = function (profile, motion) {\n      figures.forEach(function (figure) {\n        var active = !profile || figure.dataset.profile === profile;\n        figure.hidden = !active;\n        if (!(!active || !motion)) {\n          animate(figure, [{ opacity: .65, transform: 'translateY(5px)' }, { opacity: 1, transform: 'translateY(0)' }], { duration: 240, easing: 'cubic-bezier(.22,1,.36,1)' });\n          figure.querySelectorAll('.le-range').forEach(function (bar) {\n            animate(bar, [{ transform: 'scaleX(.92)' }, { transform: 'scaleX(1)' }], { duration: 420, easing: 'cubic-bezier(.22,1,.36,1)' });\n          });\n        }\n      });\n      buttons.forEach(function (button) { button.setAttribute('aria-pressed', String(button.dataset.profile === profile)); });\n      allButton.setAttribute('aria-pressed', String(!profile));\n    };\n    buttons.forEach(function (button, index) {\n      button.addEventListener('click', function () { select(button.dataset.profile, true); });\n      button.addEventListener('keydown', function (event) {\n        var next = index;\n        if (event.key === 'ArrowRight') next = (index + 1) % buttons.length;\n        else if (event.key === 'ArrowLeft') next = (index + buttons.length - 1) % buttons.length;\n        else if (event.key === 'Home') next = 0;\n        else if (event.key === 'End') next = buttons.length - 1;\n        else return;\n        event.preventDefault();\n        buttons[next].focus();\n        select(buttons[next].dataset.profile, true);\n      });\n    });\n    allButton.addEventListener('click', function () { select(null, false); });\n    select('1', false);\n    controls.hidden = false;\n  });\n  var progress = root.querySelector('.le-progress i');\n  var links = Array.from(root.querySelectorAll('.le-nav a'));\n  var targets = links.map(function (link) { return root.querySelector(link.getAttribute('href')); });\n  var scheduled = false;\n  var update = function () {\n    scheduled = false;\n    var rect = root.getBoundingClientRect();\n    var available = rect.height - window.innerHeight;\n    var fraction = available > 0 ? Math.max(0, Math.min(1, -rect.top \/ available)) : 0;\n    progress.style.transform = 'scaleX(' + fraction + ')';\n    var active = -1;\n    targets.forEach(function (section, index) {\n      if (!(!section || section.getBoundingClientRect().top > 150)) active = index;\n    });\n    links.forEach(function (link, index) {\n      if (index === active) link.setAttribute('aria-current', 'location');\n      else link.removeAttribute('aria-current');\n    });\n  };\n  var requestUpdate = function () {\n    if (!scheduled) { scheduled = true; window.requestAnimationFrame(update); }\n  };\n  window.addEventListener('scroll', requestUpdate, { passive: true });\n  window.addEventListener('resize', requestUpdate, { passive: true });\n  root.addEventListener('toggle', requestUpdate, true);\n  if ('ResizeObserver' in window) new ResizeObserver(requestUpdate).observe(root);\n  update();\n  var openedForPrint = [];\n  window.addEventListener('beforeprint', function () {\n    openedForPrint = Array.from(root.querySelectorAll('details:not([open])'));\n    openedForPrint.forEach(function (detail) { detail.open = true; });\n  });\n  window.addEventListener('afterprint', function () {\n    openedForPrint.forEach(function (detail) { detail.open = false; });\n    openedForPrint = [];\n  });\n})();\n\n(function () {\n  'use strict';\n  var root = document.getElementById('lean-etude');\n  if (!root || root.__leanUpdateScenes) return;\n\n  var hero = root.querySelector('.le-hero-scene');\n  var stage = root.querySelector('.le-time-stage');\n  var steps = stage ? Array.from(stage.querySelectorAll('.le-timeline li')) : [];\n  var day = stage ? stage.querySelector('[data-current-day]') : null;\n  var unit = stage ? stage.querySelector('[data-current-unit]') : null;\n  if (!(hero || steps.length)) return;\n\n  var reduced = window.matchMedia('(prefers-reduced-motion: reduce)');\n  var fine = window.matchMedia('(hover: hover) and (pointer: fine)');\n  var heroVisible = !!hero, stageVisible = !!stage;\n  var x = 0, y = 0, targetX = 0, targetY = 0;\n  var frame = 0, previousTime = 0, currentStep = -1;\n  var clamp = function (n, min, max) { return Math.max(min, Math.min(max, n)); };\n  var inView = function (rect) { return !(rect.bottom <= 0 || rect.top >= window.innerHeight); };\n\n  function resetHero() {\n    x = y = targetX = targetY = previousTime = 0;\n    if (!hero) return;\n    hero.style.setProperty('--pointer-x', '0');\n    hero.style.setProperty('--pointer-y', '0');\n    if (reduced.matches) hero.style.setProperty('--scene-scroll', '0');\n  }\n\n  function selectStep(index) {\n    if (index === currentStep || !steps[index]) return;\n    currentStep = index;\n    steps.forEach(function (step, i) { step.classList.toggle('is-current', i === index); });\n    stage.style.setProperty('--time-progress', String(steps.length > 1 ? index \/ (steps.length - 1) : 0));\n    var label = steps[index].querySelector('strong b') || steps[index].querySelector('strong');\n    var unitLabel = steps[index].querySelector('strong > span');\n    if (!(!day || !label)) day.textContent = label.textContent.trim();\n    if (!(!unit || !unitLabel)) unit.textContent = unitLabel.textContent.trim();\n  }\n\n  function update(time) {\n    frame = 0;\n    var continuePointer = false;\n    if (!(!hero || !heroVisible || reduced.matches)) {\n      var bounds = hero.getBoundingClientRect();\n      if (inView(bounds)) {\n        hero.style.setProperty('--scene-scroll', String(clamp(-bounds.top \/ Math.max(bounds.height, 1), 0, 1)));\n        var amount = previousTime ? 1 - Math.exp(-Math.min(time - previousTime, 50) \/ 85) : 0.2;\n        x += (targetX - x) * amount;\n        y += (targetY - y) * amount;\n        if (Math.abs(targetX - x) < 0.001) x = targetX;\n        if (Math.abs(targetY - y) < 0.001) y = targetY;\n        hero.style.setProperty('--pointer-x', x.toFixed(4));\n        hero.style.setProperty('--pointer-y', y.toFixed(4));\n        continuePointer = x !== targetX || y !== targetY;\n      } else resetHero();\n    }\n    if (!(!stage || !stageVisible || !steps.length)) {\n      var nearest = 0, distance = Infinity, midpoint = window.innerHeight * 0.5;\n      steps.forEach(function (step, index) {\n        var bounds = step.getBoundingClientRect();\n        var candidate = Math.abs(bounds.top + bounds.height * 0.5 - midpoint);\n        if (candidate < distance) { distance = candidate; nearest = index; }\n      });\n      selectStep(nearest);\n    }\n    previousTime = continuePointer ? time : 0;\n    if (continuePointer) requestUpdate();\n  }\n\n  function requestUpdate() {\n    if (!(frame || document.hidden)) frame = window.requestAnimationFrame(update);\n  }\n\n  if (hero) {\n    resetHero();\n    hero.addEventListener('pointermove', function (event) {\n      if (reduced.matches || !fine.matches || !heroVisible || event.pointerType === 'touch') return;\n      var bounds = hero.getBoundingClientRect();\n      targetX = clamp(((event.clientX - bounds.left) \/ Math.max(bounds.width, 1)) * 2 - 1, -1, 1);\n      targetY = clamp(((event.clientY - bounds.top) \/ Math.max(bounds.height, 1)) * 2 - 1, -1, 1);\n      requestUpdate();\n    }, { passive: true });\n    hero.addEventListener('pointerleave', function () {\n      targetX = targetY = 0;\n      requestUpdate();\n    }, { passive: true });\n  }\n  if (day) day.setAttribute('aria-hidden', 'true');\n  if (unit) unit.setAttribute('aria-hidden', 'true');\n  if (steps.length) selectStep(0);\n\n  if ('IntersectionObserver' in window) {\n    var observer = new IntersectionObserver(function (entries) {\n      entries.forEach(function (entry) {\n        if (entry.target === hero) {\n          heroVisible = entry.isIntersecting;\n          if (!heroVisible) resetHero();\n        }\n        if (entry.target === stage) stageVisible = entry.isIntersecting;\n      });\n      if (heroVisible || stageVisible) requestUpdate();\n    });\n    if (hero) observer.observe(hero);\n    if (stage) observer.observe(stage);\n  }\n\n  function mediaChanged() {\n    resetHero();\n    requestUpdate();\n  }\n  [reduced, fine].forEach(function (media) {\n    if (media.addEventListener) media.addEventListener('change', mediaChanged);\n    else if (media.addListener) media.addListener(mediaChanged);\n  });\n  window.addEventListener('scroll', function () {\n    if (!(!heroVisible || reduced.matches) || stageVisible) requestUpdate();\n  }, { passive: true });\n  window.addEventListener('resize', requestUpdate, { passive: true });\n  document.addEventListener('visibilitychange', function () {\n    if (document.hidden) {\n      if (frame) window.cancelAnimationFrame(frame);\n      frame = 0;\n      resetHero();\n    } else requestUpdate();\n  });\n  root.addEventListener('toggle', requestUpdate, true);\n  root.__leanUpdateScenes = requestUpdate;\n  requestUpdate();\n})();\n\n<\/script>","protected":false},"excerpt":{"rendered":"<p>On a demand\u00e9 291 fois \u00e0 ChatGPT, Gemini et Claude combien manger pour maigrir, pour 6 profils identiques. Aucune IA ne demande la masse grasse ni les pas, 97 % r\u00e9pondent par une grille d&rsquo;activit\u00e9, et le m\u00eame profil re\u00e7oit jusqu&rsquo;\u00e0 1260 kcal d&rsquo;\u00e9cart. M\u00e9thode, donn\u00e9es brutes et ce qu&rsquo;il faudrait faire \u00e0 la place.<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"single-lvm-blank","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-2163","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v25.6 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Does Losing Weight With ChatGPT Work? Study: 291 Answers, Up to 1,260 kcal of Difference - Lean<\/title>\n<meta name=\"description\" content=\"We asked ChatGPT, Gemini and Claude 291 times how much to eat to lose weight, for 6 identical profiles. No AI asks for body fat or steps, 97% answer with an activity grid, and the same profile gets up to 1,260 kcal of difference. Method, raw data and what should be done instead.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/lean-app.com\/en\/perdre-du-poids-avec-chatgpt\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Does Losing Weight With ChatGPT Work? Study: 291 Answers, Up to 1,260 kcal of Difference - Lean\" \/>\n<meta property=\"og:description\" content=\"We asked ChatGPT, Gemini and Claude 291 times how much to eat to lose weight, for 6 identical profiles. No AI asks for body fat or steps, 97% answer with an activity grid, and the same profile gets up to 1,260 kcal of difference. Method, raw data and what should be done instead.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/lean-app.com\/en\/perdre-du-poids-avec-chatgpt\/\" \/>\n<meta property=\"og:site_name\" content=\"Lean\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/share\/1GXD3qyMBy\/?mibextid=wwXIfr\" \/>\n<meta property=\"article:published_time\" content=\"2026-09-08T15:45:35+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-09-08T17:55:59+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp\" \/>\n<meta name=\"author\" content=\"L&#039;\u00e9quipe Lean\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"L&#039;\u00e9quipe Lean\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"14 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/\"},\"author\":{\"@id\":\"https:\/\/lean-app.com\/#organization\"},\"headline\":\"Perdre du poids avec ChatGPT, \u00e7a marche ? \u00c9tude : 291 r\u00e9ponses, jusqu&rsquo;\u00e0 1 260 kcal d&rsquo;\u00e9cart\",\"datePublished\":\"2026-09-08T15:45:35+00:00\",\"dateModified\":\"2026-09-08T17:55:59+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/\"},\"wordCount\":2850,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/lean-app.com\/#organization\"},\"image\":{\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp\",\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/\",\"url\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/\",\"name\":\"Does Losing Weight With ChatGPT Work? Study: 291 Answers, Up to 1,260 kcal of Difference - Lean\",\"isPartOf\":{\"@id\":\"https:\/\/lean-app.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp\",\"datePublished\":\"2026-09-08T15:45:35+00:00\",\"dateModified\":\"2026-09-08T17:55:59+00:00\",\"description\":\"We asked ChatGPT, Gemini and Claude 291 times how much to eat to lose weight, for 6 identical profiles. No AI asks for body fat or steps, 97% answer with an activity grid, and the same profile gets up to 1,260 kcal of difference. Method, raw data and what should be done instead.\",\"breadcrumb\":{\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage\",\"url\":\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp\",\"contentUrl\":\"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/lean-app.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Perdre du poids avec ChatGPT, \u00e7a marche ? \u00c9tude : 291 r\u00e9ponses, jusqu&rsquo;\u00e0 1 260 kcal d&rsquo;\u00e9cart\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/lean-app.com\/#website\",\"url\":\"https:\/\/lean-app.com\/\",\"name\":\"Lean\",\"description\":\"Download and try the app for FREE \u2014 take control of your body FOR LIFE.Download and try the app for FREE \u2014 take control of your body FOR LIFE.\",\"publisher\":{\"@id\":\"https:\/\/lean-app.com\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/lean-app.com\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/lean-app.com\/#organization\",\"name\":\"Lean\",\"url\":\"https:\/\/lean-app.com\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/lean-app.com\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/lean-app.com\/wp-content\/uploads\/2025\/08\/Logo-for-Lean-App-the-best-calorie-and-TDEE-tracking-application.png\",\"contentUrl\":\"https:\/\/lean-app.com\/wp-content\/uploads\/2025\/08\/Logo-for-Lean-App-the-best-calorie-and-TDEE-tracking-application.png\",\"width\":1672,\"height\":1484,\"caption\":\"Lean\"},\"image\":{\"@id\":\"https:\/\/lean-app.com\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.facebook.com\/share\/1GXD3qyMBy\/?mibextid=wwXIfr\",\"https:\/\/apps.apple.com\/fr\/app\/lean-calorie-ai-podometre\/id6738668646\",\"https:\/\/play.google.com\/store\/apps\/details?id=com.lean.testsqflite\",\"https:\/\/app.dealroom.co\/companies\/lean_calorie_tracker\",\"https:\/\/alternativeto.net\/software\/lean-calorie-tracker\/\"]}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Does Losing Weight With ChatGPT Work? Study: 291 Answers, Up to 1,260 kcal of Difference - Lean","description":"We asked ChatGPT, Gemini and Claude 291 times how much to eat to lose weight, for 6 identical profiles. No AI asks for body fat or steps, 97% answer with an activity grid, and the same profile gets up to 1,260 kcal of difference. Method, raw data and what should be done instead.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/lean-app.com\/en\/perdre-du-poids-avec-chatgpt\/","og_locale":"en_US","og_type":"article","og_title":"Does Losing Weight With ChatGPT Work? Study: 291 Answers, Up to 1,260 kcal of Difference - Lean","og_description":"We asked ChatGPT, Gemini and Claude 291 times how much to eat to lose weight, for 6 identical profiles. No AI asks for body fat or steps, 97% answer with an activity grid, and the same profile gets up to 1,260 kcal of difference. Method, raw data and what should be done instead.","og_url":"https:\/\/lean-app.com\/en\/perdre-du-poids-avec-chatgpt\/","og_site_name":"Lean","article_publisher":"https:\/\/www.facebook.com\/share\/1GXD3qyMBy\/?mibextid=wwXIfr","article_published_time":"2026-09-08T15:45:35+00:00","article_modified_time":"2026-09-08T17:55:59+00:00","og_image":[{"url":"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp","type":"","width":"","height":""}],"author":"L'\u00e9quipe Lean","twitter_card":"summary_large_image","twitter_misc":{"Written by":"L'\u00e9quipe Lean","Est. reading time":"14 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#article","isPartOf":{"@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/"},"author":{"@id":"https:\/\/lean-app.com\/#organization"},"headline":"Perdre du poids avec ChatGPT, \u00e7a marche ? \u00c9tude : 291 r\u00e9ponses, jusqu&rsquo;\u00e0 1 260 kcal d&rsquo;\u00e9cart","datePublished":"2026-09-08T15:45:35+00:00","dateModified":"2026-09-08T17:55:59+00:00","mainEntityOfPage":{"@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/"},"wordCount":2850,"commentCount":0,"publisher":{"@id":"https:\/\/lean-app.com\/#organization"},"image":{"@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage"},"thumbnailUrl":"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp","inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/","url":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/","name":"Does Losing Weight With ChatGPT Work? Study: 291 Answers, Up to 1,260 kcal of Difference - Lean","isPartOf":{"@id":"https:\/\/lean-app.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage"},"image":{"@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage"},"thumbnailUrl":"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp","datePublished":"2026-09-08T15:45:35+00:00","dateModified":"2026-09-08T17:55:59+00:00","description":"We asked ChatGPT, Gemini and Claude 291 times how much to eat to lose weight, for 6 identical profiles. No AI asks for body fat or steps, 97% answer with an activity grid, and the same profile gets up to 1,260 kcal of difference. Method, raw data and what should be done instead.","breadcrumb":{"@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#primaryimage","url":"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp","contentUrl":"https:\/\/lean-app.com\/wp-content\/uploads\/2026\/05\/lvm-logo-lean-square-scaled.webp"},{"@type":"BreadcrumbList","@id":"https:\/\/lean-app.com\/perdre-du-poids-avec-chatgpt\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/lean-app.com\/"},{"@type":"ListItem","position":2,"name":"Perdre du poids avec ChatGPT, \u00e7a marche ? \u00c9tude : 291 r\u00e9ponses, jusqu&rsquo;\u00e0 1 260 kcal d&rsquo;\u00e9cart"}]},{"@type":"WebSite","@id":"https:\/\/lean-app.com\/#website","url":"https:\/\/lean-app.com\/","name":"Lean","description":"Download and try the app for FREE \u2014 take control of your body FOR LIFE.Download and try the app for FREE \u2014 take control of your body FOR LIFE.","publisher":{"@id":"https:\/\/lean-app.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/lean-app.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/lean-app.com\/#organization","name":"Lean","url":"https:\/\/lean-app.com\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/lean-app.com\/#\/schema\/logo\/image\/","url":"https:\/\/lean-app.com\/wp-content\/uploads\/2025\/08\/Logo-for-Lean-App-the-best-calorie-and-TDEE-tracking-application.png","contentUrl":"https:\/\/lean-app.com\/wp-content\/uploads\/2025\/08\/Logo-for-Lean-App-the-best-calorie-and-TDEE-tracking-application.png","width":1672,"height":1484,"caption":"Lean"},"image":{"@id":"https:\/\/lean-app.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/share\/1GXD3qyMBy\/?mibextid=wwXIfr","https:\/\/apps.apple.com\/fr\/app\/lean-calorie-ai-podometre\/id6738668646","https:\/\/play.google.com\/store\/apps\/details?id=com.lean.testsqflite","https:\/\/app.dealroom.co\/companies\/lean_calorie_tracker","https:\/\/alternativeto.net\/software\/lean-calorie-tracker\/"]}]}},"_links":{"self":[{"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/posts\/2163","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/comments?post=2163"}],"version-history":[{"count":0,"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/posts\/2163\/revisions"}],"wp:attachment":[{"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/media?parent=2163"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/categories?post=2163"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lean-app.com\/en\/wp-json\/wp\/v2\/tags?post=2163"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}