Overall score
78.0
Rank #2 · A confidence
Anthropic · stable
Anthropic's most capable widely released model, documented for long-running agent workloads.
Overall score
78.0
Rank #2 · A confidence
Capability
71.6
25 eligible benchmark families
Coverage
84%
Confidence A
Benchmarks scored
30
35 raw result rows
Model specification
Identity, configuration and price are preserved as dated source records.
Category profile
A category remains Not evaluated until an eligible source-qualified result is attached.
Benchmark scores
Best published score per benchmark. Open any row for the full per-benchmark leaderboard.
| Benchmark | Category | Score | Date |
|---|---|---|---|
| τ²-Bench TelecomRanking weighted | agents | 98.538% | 2026-07-15 |
| τ-benchRanking weighted | agents | 98.5% | 2026-07-15 |
| SWE-bench VerifiedRanking weighted | coding | 95% | 2026-07-15 |
| GPQA DiamondRanking weighted | reasoning | 92.63% | 2026-07-15 |
| BenchLM Knowledge priorRanking weighted | research | 89% | 2026-07-15 |
| OSWorld-VerifiedRanking weighted | agents | 85% | 2026-07-15 |
| Terminal-BenchRanking weighted | agents | 84.64% | 2026-07-15 |
| BenchLM Math priorRanking weighted | mathematics | 83.1% | 2026-07-15 |
| BenchLM Coding priorRanking weighted | coding | 81.7% | 2026-07-15 |
| SWE-Bench ProRanking weighted | coding | 80% | 2026-07-15 |
| BenchLM Multimodal priorRanking weighted | multimodal | 79.7% | 2026-07-15 |
| LiveBenchRanking weighted | reasoning | 78.31% | 2026-07-15 |
| BenchLM Agentic priorRanking weighted | agents | 76.84% | 2026-07-15 |
| CursorBenchRanking weighted | coding | 70.5% | 2026-07-15 |
| AA Long Context ReasoningRanking weighted | research | 70% | 2026-07-15 |
| IFBench | instruction following | 63.5% | 2026-07-15 |
| IFEval | instruction following | 63.5% | 2026-07-15 |
| GDPval-AA v2Ranking weighted | agents | 62.98% | 2026-07-15 |
| Long-Horizon-Terminal-BenchRanking weighted | agents | 62.9% | 2026-07-15 |
| Terminal-Bench HardRanking weighted | agents | 62.8788% | 2026-07-15 |
| SciCodeRanking weighted | coding | 60.2% | 2026-07-15 |
| Humanity's Last ExamRanking weighted | reasoning | 53.34% | 2026-07-15 |
| GAIARanking weighted | agents | 52.3% | 2026-07-15 |
| EnterpriseOps-Gym-AARanking weighted | agents | 51.1191% | 2026-07-15 |
| AA AutomationBenchRanking weighted | agents | 48.6% | 2026-07-15 |
| AA-OmniscienceRanking weighted | research | 40.15 index | 2026-07-15 |
| Blueprint-Bench 2 | multimodal | 38.6% | 2026-07-15 |
| CritPtRanking weighted | reasoning | 28.5714% | 2026-07-15 |
| τ³-BankingRanking weighted | agents | 26.8041% | 2026-07-15 |
| Harvey LAB-AARanking weighted | agents | 14.1667% | 2026-07-15 |
Showing the best of 35 raw result rows across 30 benchmark families.
Price and efficiency
Operational values are shown only when the source publishes a comparable measurement or a valid derivation.
Verification summary
News and X posts are structurally excluded from this model's evidence graph.
Capability ranking is complete for the current snapshot. The last source check for this exact model was 2026-07-14.
Related models
Exact variants remain separate even when they share a provider or family.
Claude Opus 4 · stable
Claude Sonnet 5Claude Sonnet 5 · stable
Claude Haiku 4.5Claude Haiku 4 · stable