Claude model for demanding reasoning and long-horizon agentic work
Benchmarks
| Benchmark | Category | Score | Variant | Provenance | Source | Observed |
|---|---|---|---|---|---|---|
| AA Agentic Index | agentic | 61.3 | — | mirror | openrouter | Sep 2, 2026 |
| 61.3 | — | independent | aa | Sep 2, 2026 | ||
| AA Coding Index | coding | 81.6 | — | mirror | openrouter | Sep 2, 2026 |
| 81.6 | — | independent | aa | Sep 2, 2026 | ||
| AA Intelligence Index | intelligence | 65.7 | — | mirror | openrouter | Sep 2, 2026 |
| 65.7 | — | independent | aa | Sep 2, 2026 | ||
| Design Arena — 3d | preference | 1420 | — | mirror | openrouter | Sep 4, 2026 |
| Design Arena — codecategories | preference | 1347 | — | mirror | openrouter | Sep 4, 2026 |
| Design Arena — dataviz | preference | 1371 | — | mirror | openrouter | Sep 4, 2026 |
| Design Arena — gamedev | preference | 1425 | — | mirror | openrouter | Sep 4, 2026 |
| Design Arena — uicomponent | preference | 1343 | — | mirror | openrouter | Sep 4, 2026 |
| Design Arena — website | preference | 1330 | — | mirror | openrouter | Sep 4, 2026 |
| Median output speed | speed | 67 tok/s | — | independent | aa | Sep 4, 2026 |
| Median time to first token | speed | 273.14s | — | independent | aa | Sep 4, 2026 |
Includes index data from Artificial Analysis.
Providers
Live operational stats per hosting endpoint, via OpenRouter.
| Provider | Quant | Context | $/1M in · out | Uptime 24h | Latency | Throughput |
|---|---|---|---|---|---|---|
| Amazon Bedrock | — | 1M | $10 · $50 | — | — | — |
| Anthropic | — | 1M | $10 · $50 | 99.87% | — | — |
| Azure | — | 1M | $10 · $50 | 99.85% | — | — |
| — | 1M | $10 · $50 | 99.21% | — | — |