Vals BioMysteryBench agent leaderboard (BioMysteryBench (Vals))
We show this table for reference; we do not rank on it.
Independent agent runs on biological data investigations, split by human-solvable and human-difficult tasks.
Evaluation results published by Vals AI. Published leaderboard aggregates and evaluation settings only. Dataset contents are not included.
BioMysteryBench score on BioMysteryBench (Vals) — November 9, 2026
We mirror the published biomysterybench score view for BioMysteryBench (Vals). Claude Fable 5.2 leads the public snapshot at 83.70%, followed by Claude Sonnet 5.5 (81.11%) and GPT-6.1 Sol (79.63%). We do not use these results to rank models overall.
Claude Fable 5.2
Anthropic
Terminus 2
Terminus 2
Claude Sonnet 5.5
Anthropic
Terminus 2
Terminus 2
GPT-6.1 Sol
OpenAI
Terminus 2 · max reasoning
Terminus 2
31 modelsAgenticCurrentDisplay onlyUpdated November 9, 2026
BioMysteryBench score table (31 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Claude Fable 5.2AnthropicTerminus 2Terminus 2 | 83.70% | Not reported | Closed |
| 2 | Claude Sonnet 5.5AnthropicTerminus 2Terminus 2 | 81.11% | Not reported | Closed |
| 3 | GPT-6.1 SolOpenAITerminus 2 · max reasoningTerminus 2 | 79.63% | Not reported | Closed |
| 4 | GPT-6 AstraOpenAITerminus 2 · max reasoningTerminus 2 | 79.26% | Not reported | Closed |
| 5 | Claude Opus 5.5AnthropicTerminus 2Terminus 2 | 79.26% | Not reported | Closed |
| 6 | Claude Opus 5AnthropicTerminus 2Terminus 2 | 79.26% | Not reported | Closed |
| 7 | Gemini 4 ArgonGoogleTerminus 2 · high reasoningTerminus 2 | 76.30% | Not reported | Closed |
| 8 | Kimi K3.1 ProMoonshot AITerminus 2 · max reasoningTerminus 2 | 75.56% | Not reported | Open |
| 9 | GPT-6 SolOpenAITerminus 2 · max reasoningTerminus 2 | 74.81% | Not reported | Closed |
| 10 | Muse Wonder FullMetaTerminus 2 · max reasoningTerminus 2 | 73.33% | Not reported | Closed |
| 11 | Kimi K3Moonshot AITerminus 2 · max reasoningTerminus 2 | 72.59% | Not reported | Open |
| 12 | Grok 4.6xAITerminus 2 · high reasoningTerminus 2 | 72.22% | Not reported | Closed |
| 13 | GPT-5.6 SolOpenAITerminus 2 · max reasoningTerminus 2 | 71.11% | Not reported | Closed |
| 14 | Claude Haiku 5.5AnthropicTerminus 2Terminus 2 | 70.37% | Not reported | Closed |
| 15 | Inkling 2 SmallThinking Machines LabTerminus 2 · 0.99 reasoningTerminus 2 | 70.00% | Not reported | Closed |
| 16 | Grok 4.7xAITerminus 2 · xhigh reasoningTerminus 2 | 69.26% | Not reported | Closed |
| 17 | Hy4 previewTencentTerminus 2Terminus 2 | 69.26% | Not reported | Open |
| 18 | MiMo-V2.6-FlashXiaomiTerminus 2Terminus 2 | 69.26% | Not reported | Open |
| 19 | DeepSeek V4.1 Flash 1017DeepSeekTerminus 2 · high reasoningTerminus 2 | 67.78% | Not reported | Open |
| 20 | Step 5 PreviewStepFunTerminus 2Terminus 2 | 67.41% | Not reported | Open |
| 21 | Ember-1FireworksTerminus 2Terminus 2 | 67.04% | Not reported | Closed |
| 22 | Muse WonderMetaTerminus 2 · max reasoningTerminus 2 | 67.04% | Not reported | Closed |
| 23 | Mistral Large 4MistralTerminus 2 · high reasoningTerminus 2 | 67.04% | Not reported | Open |
| 24 | Muse Spark 1.2MetaTerminus 2 · xhigh reasoningTerminus 2 | 64.81% | Not reported | Closed |
| 25 | DeepSeek V4 Flash 0731DeepSeekTerminus 2 · high reasoningTerminus 2 | 64.44% | Not reported | Open |
| 26 | Gemini 3.8 FlashGoogleTerminus 2 · high reasoningTerminus 2 | 62.22% | Not reported | Closed |
| 27 | GPT-6 LunaOpenAITerminus 2 · max reasoningTerminus 2 | 61.48% | Not reported | Closed |
| 28 | GPT-5.6 LunaOpenAITerminus 2 · max reasoningTerminus 2 | 61.48% | Not reported | Closed |
| 29 | Gemini 3.1 Pro PreviewGoogleTerminus 2 · high reasoningTerminus 2 | 60.00% | Not reported | Unknown |
| 30 | Gemini 3.6 FlashGoogleTerminus 2 · high reasoningTerminus 2 | 58.52% | Not reported | Closed |
| 31 | Beam 1.0ReflectionTerminus 2Terminus 2 | 42.59% | Not reported | Unknown |
How BioMysteryBench is shown here
BenchLM mirrors the public Vals AI BioMysteryBench leaderboard captured from https://www.vals.ai/benchmarks/biomysterybench and updated by Vals on November 9, 2026. The snapshot preserves overall scores, uncertainty, latency, cost-per-test metadata, and task-level scores where Vals publishes them.
BioMysteryBench is display only on BenchLM. Vals proprietary or Vals-hosted aggregate views are useful context, but BenchLM does not use them as weighted ranking inputs or as a replacement for benchmark-native source records.
Snapshot
The published BioMysteryBench (Vals) snapshot places Claude Fable 5.2 first at 83.70%. The third row is 4.07 points behind. The broader top-10 range is 10.37 points, so the table still separates the published systems.
31 models have been evaluated on BioMysteryBench (Vals). The benchmark falls in the Agentic category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. BioMysteryBench (Vals) is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About BioMysteryBench (Vals)
Year
2026
Tasks
End-to-end bioinformatics investigations from raw data
Format
Average task accuracy across three runs
Difficulty
Agentic biological analysis
Vals runs Anthropic's biological-data tasks through a fixed Terminus 2 harness and reports overall, human-solvable, and human-difficult scores. Three full-benchmark runs and their uncertainty remain source-specific, display-only evidence.
Freshness and provenance
Version
BioMysteryBench (Vals) 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does BioMysteryBench (Vals) measure?
Independent agent runs on biological data investigations, split by human-solvable and human-difficult tasks.
Which model leads the published BioMysteryBench (Vals) snapshot?
Claude Fable 5.2 currently leads the published BioMysteryBench (Vals) snapshot with 83.70% biomysterybench score. BenchLM shows this benchmark for display only and does not use it in overall rankings.
How many models are evaluated on BioMysteryBench (Vals)?
The November 9, 2026 snapshot contains 31 AI models.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.