Claude Opus 5 vs DeepSeek V4 Pro 0813
| model Claude Opus 5 Anthropic | model DeepSeek V4 Pro 0813 DeepSeek | |
|---|---|---|
| Context | 1M | 1M |
| Input $/M | — | — |
| Output $/M | — | — |
| Knowledge cutoff | 2026-05 | — |
| Available via | Anthropic API | DeepSeek API |
| SWE-bench Verified (resolved) | 96 | — |
| SWE-bench Pro | 79.2% | — |
| SWE-bench Multilingual | 89.5% | — |
| SWE-bench Multimodal | 59.4% | — |
| DeepSWE | 68.8% | — |
| FrontierCode (mean@5) | 53.4 | — |