OncologyBench Leaderboard: Anthropic Sweeps Top Six as Kimi K3 Leads the Rest at 24%
kenbwork · x · 2026-10-14
The full OncologyBench leaderboard is up on benchmarks.bio. Under strict task+decision scoring, Claude Opus 5 (Pi, 40.0%), Opus 5.5 (Claude Code, 37.3%) and Opus 5 (Claude Code, 36.0%) top the table, with all top six spots going to Anthropic models. Kimi K3 is the highest non-Claude entry at 24.0%; GPT-6 variants score 18.7–21.3% and Grok 4.6/4.7 17.3–18.7%. Scores are reported per model-harness pair since the harness substantially changes results, and refusals and timeouts count in the denominator.
Related event: OncologyBench Launch: Best AI Models Pass Only 29-40% of Oncology Decisions(6 posts)→
More from Models
- Gemini triggers SAFETY refusal mid-reverse-engineering, but finishes the dump first — banteg · 2026-10-14
- Gemini 4 Argon supports up to 1M token outputs, 8x any frontier GPT or Claude model — daniel_mac8 · 2026-10-14
- Brundage spots COT-like elements leaking into Argon's non-COT replies — Miles_Brundage · 2026-10-14
- NVIDIA open-sources PersonaPlex-7B, a full-duplex voice model that listens and speaks simultaneously — solyarisoftware · 2026-10-14
- Muse Wonder Tops TaxCalcBench at 71% — and Is First Model Caught Cheating — acbuller · 2026-10-14
- 'Two-way RL' for task-level model compression: building wafer-thin hyper-specialized models — tokenbender · 2026-10-14