Anthropic
Claude 3.5 Sonnet (Oct 2024)
- Announced
- Access
- Public
Benchmark scores
| Test | Score | Best on the tracker | How it was measured |
|---|---|---|---|
| Humanity's Last Examexpert questions, no tools | 4.1% | GPT-6 Astra 55% | Independentleaderboard entry "Claude 3.5 Sonnet (October 2024)" (claude-3-5-sonnet-20241022) |
| GPQA DiamondPhD-level science questionsPhD experts: 70%Saturated | 55% | Gemini 4 Argon 96% | Independentdefault settings (claude-3-5-sonnet-20241022)Run 27 Jan 2025 |
| OTIS Mock AIMEcompetition mathsSaturated | 8.5% | GPT-5.5 100% | Independentdefault settings (claude-3-5-sonnet-20241022)Run 25 Feb 2025 |
| METR time horizontask length an AI agent finishes half the timeNo longer updated | 21 min | Claude Mythos Preview 17 h | IndependentMETR Horizon v1.1 run 'claude_3_5_sonnet_20241022_inspect'; 95% CI 10.1–40.8 min |
Also on these charts
Sources
- Epoch AI Benchmarking Hub (model release dates and runs)Epoch AI · Dataset · 9 Nov 2026 · epoch.ai