Peptides AI Visibility Index 2026

Defy Medical led the measured Peptides & Weight Loss appearance leaderboard at 12.4% of 558 answers returned, with a 95% interval of 9.9% to 15.4%. The result does not establish a clear lead over Ro, Hims, BPC-157, or Overview because those intervals overlap Defy Medical's interval.

The study tested 48 prompts across weight management, recovery and healing, longevity and anti-aging, cognitive and mood, regulatory or legal status, and clinic or provider selection. AI visibility is fragmented by engine: ChatGPT named Defy Medical in 26 of 128 answers, while Gemini named Overview in 56 of 144 answers.

The measured prompts describe this category on September 10, 2026. They do not describe all consumer questions, consumer demand, clinical outcomes, or regulatory compliance.

Which Brands Lead the AI Peptide & Weight-Loss Leaderboard?

Defy Medical ranks first with a 12.4% appearance rate. Ro ranks second at 10.9%, Hims ranks third at 10.6%, BPC-157 ranks fourth at 10.4%, and Overview ranks fifth at 10.0%. The top results mix clinics, products, a regulator, and non-brand entities.

Because these entities are not a single comparable commercial panel, the measured data does not support a composite comparison of telehealth clinics against pharmaceutical GLP-1 brands. It does show that provider, product, compound, and regulatory names all enter answers to the same prompt set.

Appearance rate is the share of the 558 returned AI answers in which each brand was named.

Defy Medical's 95% interval overlaps Ro, Hims, BPC-157, and Overview. The difference among ranks one through five is within the margin of this sample. The same caution applies throughout the leaderboard whenever intervals overlap.

The profiles below use appearance rates calculated from answers returned. Citation rates use the separate denominator of 555 answers with retrieval evidence.

Leading Brand Profiles

Where Do the AI Engines Disagree on Peptide Brands?

The engines produce different brand patterns despite a 96.4% retrieval rate across the study. Perplexity returned 144 answers, ChatGPT returned 128, Gemini returned 144, and Claude returned 142. Each engine also uses a different model and retrieval tool, so a gap can reflect system design rather than a stable difference in a brand's underlying visibility.

Each cell uses that engine's own answer count as its denominator, so engines can be compared row by row.

Gemini is the only engine that named Overview, doing so in 56 of 144 answers. ChatGPT drove Peptide Sciences Corp's result, naming it in 36 of 128 answers, versus 1 of 144 for Perplexity and 0 of 142 for Claude. ChatGPT also named Defy Medical in 26 of 128 answers and Hims in 24 of 128 answers.

Perplexity named BPC-157 in 27 of 144 answers, compared with 7 of 128 on ChatGPT and 1 of 142 on Claude. It also named Novo Nordisk in 24 of 144 answers, Ozempic in 21 of 144, Wegovy in 22 of 144, Semax in 20 of 144, Selank in 20 of 144, and CJC-1295 in 19 of 144. ChatGPT named CJC-1295 in 0 answers, while Claude also named it in 0 answers.

Claude returned a low but not uniform pattern. It named Defy Medical and Fountain Life in 7 of 142 answers each, Ro in 6, and Henry in 4. It named Novo Nordisk, Peptide Sciences Corp, Eli Lilly and Company, Cerebrolysin, Saxenda, Invigor Medical, Victoza, and Egrifta in 0 answers. Average first mention position also ranges widely among named entities, from Fountain Life at 3.1 to Egrifta at 20.1.

Which Demand Clusters Carry Each Brand?

The cluster results identify where each named entity appears within the measured question groups. The chart separates general buyer-intent questions from sub-category or specific-need questions, so cluster leadership should be read within its group rather than as one combined selection score.

The same answers regrouped by the demand cluster each prompt was sampled for; cluster rates use that cluster's own answer count.

Use the cluster pattern to separate broad category visibility from narrower needs. A brand that leads a specific-need cluster may not lead general provider-selection questions. The tally does not provide a basis to extend cluster performance beyond the measured prompts.

Which Sites Get Cited in AI Peptide Answers?

Owned-domain citation rates use 555 answers with retrieval evidence, not the 558 answers returned. The U.S. Food and Drug Administration and FDA each have an 18.9% owned-domain citation rate, based on the same fda.gov domain. Among commercial entities, Fountain Life has a 5.9% rate, followed by Defy Medical at 5.0%, Hone Health at 4.1%, and Marek Health at 3.2%.

Share of the 555 answers with retrieval evidence that cited a URL on the brand's own website.

Several visible entities had little or no owned-site citation. Peptide Sciences Corp and Wegovy each had a 0.2% owned-domain citation rate. Zepbound, Saxenda, Victoza, and Egrifta each had a 0% rate. Those results use the retrieval-evidence denominator.

Publishers cited as evidence, counted once per answer rather than once per link.

Third-party publishers carried much of the cited material. formblends.com appeared in 182 answers with retrieval evidence, a 32.8% answer penetration rate. fda.gov appeared in 104 answers, or 18.7%, followed by healingmaps.com at 98 answers and 17.7%, and mypeptidematch.com at 94 answers and 16.9%.

Who Is Winning and Who Is at Risk in Peptide AI Visibility?

Measured Winners

Defy Medical is a measured leader at rank one with a 12.4% appearance rate, a 3.8 average first mention position, and a 5.0% owned-domain citation rate among retrieval-evidenced answers. Fountain Life is lower in overall appearance at 7.0%, but it has a 3.1 average first mention position and a 5.9% owned-domain citation rate. The U.S. Food and Drug Administration and FDA have the highest measured owned-domain citation rate at 18.9%.

Ro and Hims remain prominent in the appearance tally at 10.9% and 10.6%, respectively. Both appeared across all four engines, though Claude named Ro in 6 of 142 answers and Hims in 1 of 142.

Measured Risk Signals

Overview depends entirely on Gemini, which named it in 56 of 144 answers while the other three engines named it in 0 answers. Peptide Sciences Corp has an 8.6% overall appearance rate but a 0.2% owned-domain citation rate and 0 Claude mentions. Novo Nordisk had 0 Claude mentions and a 2.0% owned-domain citation rate despite appearing in 8.4% of returned answers.

Zepbound appeared in 5.0% of returned answers and had a 0% owned-domain citation rate among answers with retrieval evidence. Saxenda, Victoza, and Egrifta also had 0% owned-domain citation rates. EVEXIAS Health Sciences had a 0% appearance rate, with a 95% interval of 0% to 0.7%, while its owned-domain citation rate was 0.4%.

What Structural Patterns Does the Tally Reveal?

The top of the appearance leaderboard is tightly clustered. Defy Medical's 9.9% to 15.4% interval overlaps the intervals for Ro, Hims, BPC-157, and Overview. The sample therefore cannot distinguish a clear first-place leader among those five entities.

The category includes strong engine concentration. Overview appeared only in Gemini. Peptide Sciences Corp was most concentrated in ChatGPT, while CJC-1295 appeared in Perplexity and Gemini but not ChatGPT or Claude. Claude produced 142 answers, yet it named several high-ranking category entities zero times.

Citation evidence also leans toward third-party publishers. formblends.com had a 32.8% answer penetration rate among answers with retrieval evidence, while the highest measured owned-domain rate was 18.9% for fda.gov. Named entities from the brief appear near zero in places: EVEXIAS Health Sciences had 0% appearance, while Egrifta had 1.3% and Victoza had 1.6%.

What Changed During the Measurement Window?

This study provides one measurement window, not a prior wave comparison. It cannot identify changes over time. It does document cross-engine divergence, third-party publisher concentration, and the separate treatment of Novo Nordisk, Ozempic, Wegovy, Eli Lilly and Company, Zepbound, and Mounjaro in the tally.

What Should Brands Do With These Measured Gaps?

• Audit engine-specific naming. Overview appeared in Gemini only, and Peptide Sciences Corp appeared in 36 of 128 ChatGPT answers but 0 Claude answers. Re-measure appearance by engine after publishing pages that use the brand's canonical name consistently.
• Build citeable owned pages for products with no owned-domain citations. Zepbound, Saxenda, Victoza, and Egrifta each recorded 0% owned-domain citation rates among 555 retrieval-evidenced answers. Re-measure owned-domain citation rate against the same denominator.
• Review answer placement for brands named late. Egrifta's average first mention position was 20.1, Victoza's was 14.9, and Saxenda's was 12.4. Re-measure first mention position after creating direct pages for the measured question types.
• Separate provider-selection content from compound and product content. The tally mixes clinics, compounds, products, and the U.S. Food and Drug Administration. Re-measure cluster results after each content type has a distinct page and clear entity naming.
• Check whether third-party source dependence is displacing owned material. formblends.com appeared in 182 retrieval-evidenced answers, while several branded sites had rates below 1%. Re-measure citations after publishing factual, source-linked pages that address the measured concerns.
• Test regulatory pages in every engine. FDA and the U.S. Food and Drug Administration each had an 18.9% owned-domain citation rate, but their appearance varied across engines. Re-measure regulatory and legal-status prompts separately after revising pages that cite the regulator directly.

What Does This Mean for Brands in the Peptide & Weight-Loss Category?

Interpretation: appearance, citation, and answer position are separate measures. A brand can appear often without receiving many owned-site citations, as Peptide Sciences Corp does. A brand can appear less often but enter answers earlier and receive more owned-site citations, as Fountain Life does.

Interpretation: brand plans need engine-level measurement. Gemini, ChatGPT, Perplexity, and Claude produced different naming patterns across the same study. A single blended rate can hide a complete absence on one engine.

How Was the Study Measured?

This flagship study tested 48 stratified prompts three times per engine across Perplexity, ChatGPT, Gemini, and Claude. It attempted 576 answers, returned 558 answers for appearance measurement, and found retrieval evidence in 555 answers for citation measurement. The data records mentions and cited sources, not medical quality, treatment safety, consumer demand, or regulatory compliance.

How This Study Was Measured

• Sampling frame fixed first: 48 prompts allocated across the demand clusters before anything was asked.
• Same questions put to every engine: Perplexity (sonar-pro), ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic), 3 trial(s) each.
• 558 answers returned of 576 attempted, 555 with retrieval evidence.
• Product lines rolled up to parent brands using a registry fixed before scoring.
• Appearance rate reported with a 95% Wilson interval; citation metrics counted only on answers with evidence.

Appearance rates use answers returned. Citation and source rates use answers with retrieval evidence. The prompt set is limited to the measured categories and date, and overlapping intervals limit ranking claims near the top of the leaderboard.

Methodology Appendix

This appendix is generated directly from the measurement record for this study. It is included so any brand named here can reproduce or challenge the result.

What Was Measured

• Category: Peptides & Weight Loss
• Measurement date: 2026-09-10
• Depth: flagship — 48 prompts, 3 trial(s) per prompt per engine, 4 engines
• Answers attempted: 576
• Answers returned (appearance denominator): 558
• Answers with retrieval evidence (citation denominator): 555 (96.4% of attempted)
• Answers that failed outright: 10
• Analysis version: study-methodology-v2 · entity registry: entity-registry-2026-08-27

Claim limit: Flagship depth: 24 stratified prompts measured three times per engine, so both between-prompt and repeat-generation variation are observable.

How the Prompts Were Sampled

Fixed sampling frame: 44 buyer-intent need(s) from the client's brief are asked verbatim, one prompt each, grouped under: Weight Management / Recovery & Healing / Longevity & Anti-Aging / Cognitive & Mood / Regulatory / Legal Status; then (the prompt budget of 48 could not give a dedicated cell to Weight Management, Recovery & Healing, Longevity & Anti-Aging, Cognitive & Mood, Regulatory / Legal Status, Clinic / Provider Selection, so that demand is covered inside the generic clusters instead), then a documented set of generic commercial demand clusters (general best, everyday mainstream choice, price/value, primary use case, beginner, retail channel, gifting/occasion, specific need, comparison, trust, ownership attribution, quality tier, durability, specialist, emerging) receives the remaining allocation round-robin. The prompt writer supplies wording inside these cells only and does not choose the category's demand structure.

Results are not strictly comparable across engines: each engine uses a different model, a different retrieval tool and a different answer format, so a gap between two engines can reflect system design as much as brand visibility.

How Each Metric Is Defined

• Appearance rate — the share of returned answers that named the brand. Measured on every answer the engine returned, whether or not it searched the web, so an engine's retrieval behaviour does not silently change the denominator.
• Macro appearance rate — the mean of the per-prompt appearance rates, so each prompt counts once regardless of how many engines answered it.
• 95% interval — Wilson score interval on the appearance rate. Where two brands' intervals overlap, this study does not support a claim that one is more visible than the other.
• Owned-domain citation rate — the share of answers with retrieval evidence that cited a URL on the brand's own domain. This is a narrow measure of the brand's own site being used as a source; it is not a measure of total mentions.
• Share of all citations — the brand's own domain as a share of all attributed publisher citations in the study.
• Not measured — the brand's official domain could not be established, so no citation result exists. This is never reported as 0%.
• 0% — the metric could be evaluated and the event did not occur.
• Average position — mean position of the brand's first mention in the answer's ordered list of named brands.
• Cross-engine consistency — the share of engines that named the brand at least once.
• Repeat consistency — the share of repeated prompt/engine cells where the brand appeared in every trial. Lower values mean the engine's answer is unstable rather than the brand being absent.

Evidence Rules

• An answer counts as having retrieval evidence only when the engine returned a retrieval receipt (search annotations or grounding records). A URL the model merely typed into its prose is not treated as evidence that it searched.
• Sources are attributed to the publisher behind each retrieval receipt. Intermediary and redirect hosts are never reported as publishers.
• Publisher identity could not be established for 5 retrieval receipt(s); those are disclosed here rather than attributed or dropped.
• Source counts are answer-level: a publisher counts once per answer in which it was cited, not once per URL.

Entity Handling

Product lines and model names roll up to their parent brand for brand-level metrics, using a registry fixed before scoring. Nothing is merged on the basis of this study's own counts alone.

No roll-ups were applied in this study.

40 further name(s) picked up in the answers were not confidently resolvable to a brand in this category and are held for human classification before the next study; they carry no metrics here.

Limits of These Findings

• This is a point-in-time measurement of what the named systems returned for these exact prompts on 2026-09-10. Engine behaviour changes without notice.
• The prompt set is a stratified sample of category demand, not a census. Results describe these prompts and should not be read as representative of all searches in the category.
• Where intervals overlap, differences between brands are not statistically distinguishable at this sample size.
• A brand not listed here was not measured as absent — it simply was not named in these answers.
• Appearance is a visibility measure, not an endorsement, a quality judgement or a market-share estimate.
• 5 of 32 ranked brand(s) had no official domain established before scoring, so they carry appearance results only and are excluded from citation metrics and source tables rather than being reported as zero.