AnthropicClaude 3 Opus
Best configuration ranks #539 of 610 on the BenchLeader Index at 38.9 ±4.2. Last measured 2 Sept 2026. Released 4 Mar 2024.
- Blended price
- $30.00/M
- $15.00 in · $75.00 out
- Output speed
- –
- First answer
- –
- Context
- 200k
- Overall index39
- Reasoning36
- Coding38
- Agents & tools38
- Maths33
- Knowledge27
- Human preference49
- Multimodal31
- Composite40
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| GPQA Diamond | 47.2%#213 | Epoch AI Benchmarking Hub |
| SimpleBench | 23.5%#73 | SimpleBench |
| LMArena Hard Prompts | 1329#209 | LMArena |
| GPQA Diamond (AA)not in index | 48.9%#412 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 2.8%#519 | Artificial Analysis |
Coding
| Benchmark | default | Source |
|---|---|---|
| WeirdML | 19.2%#136 | WeirdML |
| LMArena Coding | 1355#217 | LMArena |
| SWE-bench Verified (any scaffold)not in index | 15.8%#63 | SWE-bench |
Maths
| Benchmark | default | Source |
|---|---|---|
| OTIS Mock AIME | 4.7%#226 | Epoch AI Benchmarking Hub |
| MATH Level 5 | 37.5%#63 | Epoch AI Benchmarking Hub |
Knowledge
| Benchmark | default | Source |
|---|---|---|
| SimpleQA Verified | 12.6%#70 | Epoch AI Benchmarking Hub |
Human preference
| Benchmark | default | Source |
|---|---|---|
| LMArena Text | 1322#211 | LMArena |
Multimodal
| Benchmark | default | Source |
|---|---|---|
| LMArena Vision | 1023#118 | LMArena |
| VISTA | 27.8%#50 | Scale AI SEAL |
| MMMU (validation) | 59.4%#33 | MMMU |
Composite
| Benchmark | default | Source |
|---|---|---|
| Epoch Capabilities Indexnot in index | 126.9#146 | Epoch AI Benchmarking Hub |
| AA Intelligence Index | 8.7#355 | Artificial Analysis |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.029 | – | – |
| Summarise a 30-page report | 12,000 / 600 | $0.225 | – | – |
| Code edit | 6,000 / 1,500 | $0.203 | – | – |
| Agentic coding session | 60,000 / 4,000 | $1.20 | – | – |
| Structured extraction | 2,000 / 200 | $0.045 | – | – |
See also
Data as of 9 Sept 2026. Compare with another model.