Anthropic: Claude Sonnet 4.5 Benchmarks
Try Claude Sonnet 4.5 on us. Starter credit covers your first requests, then one key reaches this route and every other model in the catalog.
anthropic/claude-sonnet-4.5
Published benchmark scores
Benchmark scores for Anthropic: Claude Sonnet 4.5 — every row links to its source, and a score is only ever attached to the exact checkpoint it was measured on. Vendor model-card and open-leaderboard numbers are cited, not run by us. Rows marked TrustedRouter · replays published are our own runs of this model through the gateway, with the full per-item replay published in trustedrouter-benchmarks so anyone can re-grade them.
| Benchmark | Category | Score | Source |
|---|---|---|---|
| OSWorld computer use |
Agentic | 61.4% | Anthropic — Claude Sonnet 4.5 2025-09-29 |
| SWE-bench Verified avg of 10 trials, 200K thinking budget; no test-time compute |
Coding | 77.2% | Anthropic — Claude Sonnet 4.5 2025-09-29 |
TrustedRouter measurements
TrustedRouter publishes route and status measurements without storing prompt or output content. Provider latency and uptime are exposed through the model performance and uptime pages.
External benchmark references
- TrustedRouter performance pageTrustedRouter measurement
- TrustedRouter uptime pageTrustedRouter measurement
- Anthropic model docsOfficial model information
- LMArena leaderboardIndependent benchmark index
- LiveBenchIndependent benchmark index
- Artificial Analysis modelsIndependent benchmark index
- HELMIndependent benchmark index