Claude Opus 5
Claude Opus 5 is Anthropic’s flagship model for demanding reasoning, coding, and long-horizon agentic work. It is particularly strong at end-to-end software tasks, code review and bug finding, visual analysis...
Everything else
| Max output | 128K |
|---|---|
| Input / 1M | $5 |
| Output / 1M | $25 |
| Cache read / 1M | $0.5 |
| Cache write / 1M | $6.25 |
| 1M in + 1M out | $30 |
| Takes | text, image, file |
| Returns | text |
| Tools | yes |
| Reasoning | optional · max, xhigh, high, medium, low |
| Open weights | no |
| Released | 2026-07-24 |
| Knowledge cutoff | — |
| Retires | — |
| Catalog id | anthropic/claude-opus-5 |
Benchmarks
| source | epoch |
|---|---|
| eci | 159.38 |
| confidence | Confident |
| accessibility | API access |
| gpqa | 0.939 |
| gpqaStderr | 0.0148 |
| aime | 0.989 |
| aimeStderr | 0.0111 |
| simpleQA | 0.567 |
| simpleQAStderr | 0.0157 |
Benchmark scores by Epoch AI, CC BY 4.0.
Using it
In 00, this model is picked per agent — and per tier, so the model that answers your customers need not be the one that writes your code. Get 00.