Model
Claude Opus 5.5
Anthropic · closed weights
Knowledge & Reasoning
Knowledge & Reasoning| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| CritPt | 31.7 | accuracy | N/A | Independent | low | epoch.ai |
| GPQA Diamond | 90.6 | accuracy | N/A | Independent | high | epoch.ai |
| SimpleQA Verified | 72.2 | accuracy | N/A | Independent | high | epoch.ai |
| TasteVal | 2.3 | compute_multiplier | N/A | Independent | medium | arxiv.org |
Coding
Coding| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| PremiseSQL | 0.2 | ex | N/A | Independent | medium | arxiv.org |
| SciCode | 66.9 | score | N/A | Independent | medium | artificialanalysis.ai |
| Zero2Repo | 81.8 | success_rate | claude code | Independent | medium | arxiv.org |
Agentic / Tool Use
Agentic / Tool Use| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| AgentPersonaBench | 76.6 | accuracy | N/A | Independent | medium | arxiv.org |
| Argo-Bench | 59.5 | score | N/A | Independent | medium | arxiv.org |
| Autonomous Science Bench | 67.0 | success_rate | N/A | Independent | medium | arxiv.org |
| BioStudyBench | 93.3 | pass_rate | N/A | Independent | medium | arxiv.org |
| BrickBench | 1249.0 | elo | N/A | Independent | medium | arxiv.org |
| CineRecon | 76.5 | score | claude code 2.1.280 | Independent | medium | arxiv.org |
| DAYJOB | 24.7 | success_rate | N/A | Independent | medium | arxiv.org |
| FDE-Gym | 0.6 | score | N/A | Independent | medium | arxiv.org |
| Incident-Arena | 0.5 | pass@1 | claude code | Independent | medium | arxiv.org |
| KlinikeBench | 28.8 | strict_pass@1 | N/A | Independent | medium | arxiv.org |
| Kolab | 0.7 | f1 | N/A | Independent | medium | arxiv.org |
| Mine Odyssey | 73.9 | success_rate | N/A | Independent | medium | arxiv.org |
| OfficeTown Bench | 50.7 | pass@1 | N/A | Independent | medium | arxiv.org |
| PersonaCUABench | 18.6 | success_rate | N/A | Independent | medium | arxiv.org |
| RoboQuest | 13.8 | success_rate | N/A | Independent | medium | arxiv.org |
| RobotWorld | 15.5 | success_rate | N/A | Independent | medium | arxiv.org |
| ToolBottleneckBench | 97.2 | success_rate | toolbottleneckbench | Independent | medium | arxiv.org |
Multimodal
Multimodal| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| CREATE | 24.9 | accuracy | N/A | Independent | medium | arxiv.org |
| GDP.pdf | 30.6 | gdp.pdf_score | N/A | Independent | medium | surgehq.ai |
| HSS | 44.6 | pass@1 | N/A | Independent | medium | arxiv.org |
| PhysVista | 90.7 | order_accuracy | N/A | Independent | medium | arxiv.org |
| VELD-QA | 74.9 | accuracy | top-3 rag (colpali) | Independent | medium | arxiv.org |
Long Context
Long ContextInstruction Following
Instruction Following| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| Design Creativity Bench | 51.0 | overall | N/A | Independent | medium | arxiv.org |
| RealCompanion | 0.7 | f1 | N/A | Independent | medium | arxiv.org |
Safety & Alignment
Safety & Alignment| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| BrowseSafe++ | 86.5 | f1 | N/A | Independent | medium | arxiv.org |
| CWE-bench | 67.0 | pass@1 | claude code | Independent | medium | arxiv.org |
Abstract Reasoning
Abstract ReasoningReproduced by the community
Loading community runs…