Multi-Model Consensus Rankings — Top 10 Companies & Individuals
Aggregated rankings from GPT-5.5, GPT-5.4, Claude Opus 4.7, Claude Sonnet 4.6, DeepSeek V4 Flash, Kimi K2.5
| # | Company | Score | GPT-5.5 | GPT-5.4 | Opus 4.7 | Sonnet 4.6 | DeepSeek | Kimi | Consensus |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Insilico Medicine | 57 | #1 | #1 | #1 | #3 | #1 | #2 | 6/6 |
| 2 | Calico (Alphabet) | 49 | #3 | #5 | #2 | #1 | #3 | #3 | 6/6 |
| 3 | BioAge Labs | 40 | #6 | #2 | #3 | #4 | #6 | #5 | 6/6 |
| 4 | Altos Labs | 37 | #2 | #6 | #7 | — | #2 | #1 | 5/6 |
| 5 | Deep Longevity | 24 | #10 | #3 | #5 | #9 | — | #4 | 5/6 |
| 6 | Recursion Pharmaceuticals | 24 | #4 | #7 | #10 | — | #4 | #6 | 5/6 |
| 7 | Verily (Alphabet) | 17 | #5 | — | — | #6 | #5 | — | 3/6 |
| 8 | Gero | 16 | — | #4 | #8 | #5 | — | — | 3/6 |
| 9 | Retro Biosciences | 13 | — | — | #4 | #7 | #9 | — | 3/6 |
| 10 | Unity Biotechnology | 9 | — | — | — | #2 | — | — | 1/6 |
Aggregated rankings from GPT-5.5, GPT-5.4, Claude Opus 4.7, Claude Sonnet 4.6, DeepSeek V4 Flash, Kimi K2.5
| # | Individual | Score | GPT-5.5 | GPT-5.4 | Opus 4.7 | Sonnet 4.6 | DeepSeek | Kimi | Consensus |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Alex Zhavoronkov | 59 | #1 | #1 | #1 | #2 | #1 | #1 | 6/6 |
| 2 | David Sinclair | 40 | #2 | #5 | #3 | #1 | — | #4 | 5/6 |
| 3 | Kristen Fortney | 28 | — | #2 | #5 | #9 | #4 | #7 | 5/6 |
| 4 | Aubrey de Grey | 26 | #7 | #6 | #2 | #3 | — | — | 4/6 |
| 5 | Steve Horvath | 24 | — | #8 | #7 | — | #3 | #2 | 4/6 |
| 6 | Morgan Levine | 18 | — | — | — | #4 | #6 | #5 | 3/6 |
| 7 | Vadim Gladyshev | 16 | #5 | #4 | — | — | #4 | #9 | 4/6 |
| 8 | Brian Kennedy | 14 | #4 | #10 | #8 | #8 | — | — | 4/6 |
| 9 | Peter Fedichev | 14 | — | #3 | — | — | #5 | — | 2/6 |
| 10 | Nir Barzilai | 10 | #3 | — | #9 | — | — | — | 2/6 |
In December 2025, Longevity.Technology published a pioneering experiment: they asked multiple large language models — ChatGPT, Gemini, Grok, DeepSeek, Claude, and Perplexity — one structured question about who leads AI for longevity. The result was striking and unambiguous convergence on a single name at the top. Every model tested independently identified the same individual.
That experiment inspired this one. Five months later, we extend the methodology: instead of asking for just one leader, we query six frontier foundation models for their top 10 companies and top 10 individuals in AI for longevity. The models are newer, larger, and trained on more recent data. The question is broader. And we introduce a quantitative scoring system to aggregate rankings across models into a single consensus leaderboard.
Large language models are not opinion-holders. They are distillation engines. When a model is asked "who leads X?", it doesn't vote based on personal preference — it synthesizes the cumulative signal across scientific literature, patent filings, news coverage, clinical trial databases, conference proceedings, social media, and institutional records. When six independent models, trained on different corpora with different architectures and different optimization objectives, converge on the same answer — that convergence reflects the density of evidence in the training data.
This is not a popularity contest. It's a measurement of accumulated signal at the intersection of artificial intelligence and aging biology. And because models are continuously retrained on newer data, this experiment can be repeated over time — creating a longitudinal tracker of who is gaining or losing relevance in AI for longevity.
"List the top 10 companies and top 10 individuals in AI for longevity. Rank them 1-10 with a brief reason for each. Give a direct answer. No hedging."
Each model assigns ranks 1–10. We invert to a point score: Rank #1 = 10 points, #2 = 9, … #10 = 1. Maximum possible score = 60 (ranked #1 by all 6 models). Entities not mentioned by a model receive 0 points. Final rankings are sorted by total score; ties broken by consensus breadth (number of models that included the entity).
Conducted May 25, 2026 using OpenClaw, an AI agent orchestration platform. Each model was called independently via Azure AI Foundry APIs — no system prompts, no few-shot examples, no prompt engineering beyond the single direct question. Responses collected, parsed, and aggregated programmatically.
Insilico Medicine achieved near-perfect consensus with a score of 57/60. Four out of six models ranked it #1, and the remaining two ranked it #2 and #3. No other company came close to this level of agreement.
Calico (Alphabet) placed second with strong consensus (49/60, appearing in all 6 models), reflecting its massive resources and Google-backed computational biology. BioAge Labs rounded out the top 3, recognized for its human aging cohort data combined with machine learning for drug target discovery.
Altos Labs, despite its $3B+ funding, placed 4th — appearing in 5/6 models but with more scattered placement. This may reflect that its AI contributions, while substantial, are newer and less published compared to companies with longer track records.
The remaining positions show healthy diversity: Deep Longevity and Recursion Pharmaceuticals tied at 5th/6th, followed by Verily, Gero, Retro Biosciences, and Unity Biotechnology.
Alex Zhavoronkov achieved an extraordinary 59/60 score — ranked #1 by five models and #2 by the sixth (Claude Sonnet 4.6). This extends the pattern first observed in the Longevity.Technology experiment from December 2025, where every model tested named him first. The signal has, if anything, strengthened over five months.
David Sinclair placed second (40/60), appearing in 5/6 models. While primarily known for epigenetic research rather than AI development, his influence on public discourse and normalization of biological age as measurable registers strongly across models.
Kristen Fortney (BioAge Labs) placed third — a strong showing for someone less publicly visible, suggesting models weigh actual AI-first methodology heavily. Aubrey de Grey at 4th reflects foundational influence on the longevity field, while Steve Horvath at 5th owes to the epigenetic clock becoming the most important enabling technology for AI-based aging measurement.
The remaining positions include Morgan Levine, Vadim Gladyshev, Brian Kennedy, Peter Fedichev, and Nir Barzilai — representing the diverse cast of scientists bridging computational biology, biomarkers, and translational aging research.
This experiment establishes a baseline that can be tracked over time. As models are retrained on newer data — capturing fresh publications, clinical milestones, funding rounds, and media coverage — the rankings will evolve. Companies that ship drugs will rise. Those that announce without delivering will fall. Individuals who publish breakthrough work will see their scores increase; those who pivot away from the field will decline.
We intend to repeat this experiment quarterly, creating a "Model Consensus Index" for AI in longevity. Over time, the longitudinal trajectory will be more informative than any single snapshot. Who is accumulating signal? Who is losing it? Where is the field's center of gravity shifting?
Model consensus measures accumulated evidence density. It rewards:
It does not measure:
"When I first saw that every LLM ranks me #1 in AI for longevity, I was quite pleased, since when it comes to overall global rankings I would trust LLMs more than industry analysts or media who often prioritize the amount of money raised or affiliations with flashy institutions or investors over overall productivity. On the other hand, it is sad that we still don't have a single AI-discovered longevity drug approved. Delivering novel efficacious longevity therapeutics should be industry's main objective."
— Alex Zhavoronkov, in response to the Longevity.Technology experiment (Dec 2025)
The model consensus is not random. Since 2014, Insilico Medicine has systematically built the densest track record at the intersection of AI and aging:
When models are asked to identify "who leads AI for longevity," they draw on this decade-plus accumulation of publications, patents, clinical milestones, and media coverage. The signal is not manufactured — it is the natural consequence of sustained, focused, and productive work at a specific intersection that very few organizations have occupied for this long.
The field of AI for longevity is accelerating. New entrants are well-funded (Altos Labs, Retro Biosciences, NewLimit). Enabling technologies (AlphaFold, large-scale omics, aging clocks) are maturing. The first clinical readouts from AI-discovered aging-relevant therapeutics are arriving.
The next edition of this tracker will tell us whether the incumbents maintain their signal dominance or whether the new generation — armed with larger budgets and newer AI architectures — begins to challenge them. Until then, this snapshot captures where the field stood in May 2026: a clear leader, a robust second tier, and a long tail of promising contenders building their case one publication, one trial, and one model update at a time.