LLM Stats (LifeSci Diamond (2026-11)): leaderboard
Strict pass rate over 55 regulatory affairs tasks in Arca’s November 2026 LifeSci Diamond cohort. Eighteen plain APIs use a shared Vercel AI Gateway reference harness with Exa web search, up to 40 model steps and provider defaults. Every rubric requirement must pass with no fail line; one answer per task is graded by an LLM jury. Results combine September 25–November 3 answer batches, rather than a fresh November rerun. Gateway API aliases do not identify immutable snapshots; effective reasoning, temperature and token settings were not recorded. Eleven answer keys lack expert review. Arca’s separate workflows use different tools and are measured separately.
Metric: Strict pass rate (%). Source: llm-stats.com. Saturation forecast: Around 2029. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Muse Wonder | 27.3 | #18 |
| 2 | GPT-6 Astra | 23.6 | #6 |
| 3 | GPT-6.1 Sol | 21.8 | #7 |
| 4 | Claude Opus 5 | 18.2 | #16 |
| 5 | Claude Fable 5.1 | 14.5 | #10 |
| 6 | Claude Fable 5 | 12.7 | #20 |
| 7 | Claude Opus 5.5 | 12.7 | #3 |
| 8 | Muse Spark 1.2 | 12.7 | #69 |
| 9 | Grok 4.6 | 10.9 | #29 |
| 10 | GPT-6.1 Terra | 10.9 | #15 |
| 11 | GPT-5.6 Sol | 9.1 | #22 |
| 12 | Kimi K3 | 9.1 | #35 |
| 13 | GPT-5.6 Terra | 9.1 | #40 |
| 14 | Claude Sonnet 5 | 7.3 | #65 |
| 15 | Claude Opus 4.8 | 5.5 | #39 |
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-lifesci-diamond-2026-11 · How It Works · Data refreshed daily, snapshot 2026-11-05.