Overall score
59.6
Rank #24 · A confidence
Anthropic · stable
Anthropic's Claude Sonnet 4.6 model tracked via the BenchLM July 2026 leaderboard for capability comparison.
Overall score
59.6
Rank #24 · A confidence
Capability
62.1
27 eligible benchmark families
Coverage
86%
Confidence A
Benchmarks scored
33
33 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 |
|---|---|---|---|
| SuperGPQARanking weighted | reasoning | 95% | 2026-07-15 |
| GPQA DiamondRanking weighted | reasoning | 89.9% | 2026-07-15 |
| BenchLM Multimodal priorRanking weighted | multimodal | 84.4% | 2026-07-15 |
| BenchLM Knowledge priorRanking weighted | research | 83.6% | 2026-07-15 |
| BenchLM Reasoning priorRanking weighted | reasoning | 80.9% | 2026-07-15 |
| SWE-bench VerifiedRanking weighted | coding | 79.6% | 2026-07-15 |
| τ-benchRanking weighted | agents | 79.5% | 2026-07-15 |
| MMLU-ProRanking weighted | research | 79.2% | 2026-07-15 |
| CharXivRanking weighted | multimodal | 77.4% | 2026-07-15 |
| τ²-Bench TelecomRanking weighted | agents | 75.731% | 2026-07-15 |
| LiveBenchRanking weighted | reasoning | 75.47% | 2026-07-15 |
| OSWorld-VerifiedRanking weighted | agents | 72.1% | 2026-07-15 |
| AA Long Context ReasoningRanking weighted | research | 70.6667% | 2026-07-15 |
| SWE-RebenchRanking weighted | coding | 60.7% | 2026-07-15 |
| BenchLM Coding priorRanking weighted | coding | 59.97% | 2026-07-15 |
| BenchLM Agentic priorRanking weighted | agents | 53.31% | 2026-07-15 |
| Terminal-Bench HardRanking weighted | agents | 53.0303% | 2026-07-15 |
| BenchLM Math priorRanking weighted | mathematics | 49.1% | 2026-07-15 |
| Humanity's Last ExamRanking weighted | reasoning | 49% | 2026-07-15 |
| SciCodeRanking weighted | coding | 46.9% | 2026-07-15 |
| Long-Horizon-Terminal-BenchRanking weighted | agents | 46.2% | 2026-07-15 |
| GAIARanking weighted | agents | 45.5% | 2026-07-15 |
| GDPval-AA v2Ranking weighted | agents | 43.858% | 2026-07-15 |
| IFBench | instruction following | 41.2% | 2026-07-15 |
| IFEval | instruction following | 41.2% | 2026-07-15 |
| ITBench-AARanking weighted | agents | 39.8023% | 2026-07-15 |
| FrontierMathRanking weighted | mathematics | 32.4% | 2026-07-15 |
| τ³-BankingRanking weighted | agents | 30.5155% | 2026-07-15 |
| APEX-Agents-AARanking weighted | agents | 28.0236% | 2026-07-15 |
| AA-OmniscienceRanking weighted | research | 12.3667 index | 2026-07-15 |
| OSWorldRanking weighted | agents | 8.3% | 2026-07-15 |
| Harvey LAB-AARanking weighted | agents | 4.1667% | 2026-07-15 |
| CritPtRanking weighted | reasoning | 3.1429% | 2026-07-15 |
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.
Missing publication dimensions: efficiency, value. The last source check for this exact model was 2026-07-16.
Related models
Exact variants remain separate even when they share a provider or family.
Claude 5 · stable
Claude Opus 4.8Claude Opus 4 · stable
Claude Sonnet 5Claude Sonnet 5 · stable