Claude Opus 5 vs GPT-5.6 Sol
MultiLangBench?lab
Opus 5—
GPT-5.6 Sol26.8
MirrorCode?
Opus 5—
GPT-5.6 Solat high20
| Benchmark | Opus 5 | GPT-5.6 Sol |
|---|---|---|
| Terminal-Bench 4.0 | 34.3 | 14.6 |
| DeepSWE 1.1 | 68.9 | 61.1 |
| SWE-bench Pro (v2 full) | 99.4 · xhigh | 95.5 · xhigh |
| SWE-bench Pro (v2 hard) | 98 · xhigh | 82.4 · xhigh |
| SciCode | 51.5 | 57.4 |
| Terminal-Bench Science 0.1 | 28.6 · max | 22.4 · max |
| WebDev Arena | 1,689 · max | — |
| Terminal-Bench Hard (version unstated) | — | 62.9 |
| MultiLangBenchlab | — | 26.8 |
| MirrorCode | — | 20 · high |
Specs
| Spec | Opus 5 | GPT-5.6 Sol |
|---|---|---|
| Input, per 1M | $5 | $4 |
| Output, per 1M | $25 | $20 |
| Cost per task? | $2.19 | $0.50 |
| Speed, tok/s | 55 | 85 |
| Context | 1M | 1.05M |
| Weights | Closed | Closed |
| Released | Jul 2026 | Jul 2026 |
| See images | ✓ | ✓ |
| Read PDFs | Partial | Partial |
| Hear audio | — | — |
| Watch video | — | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | Partial |
| Return strict JSON | ✓ | ✓ |
| Search the web | ? | ? |
| Run code | ? | ? |
| Use a computer | ? | ? |
What breaks when you switch
Opus 5 → GPT-5.6 Sol
Nothing known.
GPT-5.6 Sol → Opus 5