Gemini 3.8 Flash vs GPT-5.6 Sol
MultiLangBench?lab
Gemini 3.8 Flash—
GPT-5.6 Sol26.8
MirrorCode?
Gemini 3.8 Flash—
GPT-5.6 Solat high20
| Benchmark | Gemini 3.8 Flash | GPT-5.6 Sol |
|---|---|---|
| Terminal-Bench 4.0 | 19.7 | 14.6 |
| DeepSWE 1.1 | 71 | 61.1 |
| SWE-bench Pro (v2 full) | 94.9 · high | 95.5 · xhigh |
| SWE-bench Pro (v2 hard) | 58.8 · high | 82.4 · xhigh |
| SciCode | 55.1 | 57.4 |
| Terminal-Bench Science 0.1 | 10.5 | 22.4 · max |
| FrontierCode 1.1 | 37.1 | — |
| WebDev Arena | 1,585 · high | — |
| Terminal-Bench Hard (version unstated) | — | 62.9 |
| MultiLangBenchlab | — | 26.8 |
| MirrorCode | — | 20 · high |
Specs
| Spec | Gemini 3.8 Flash | GPT-5.6 Sol |
|---|---|---|
| Input, per 1M | !$0.75 | $4 |
| Output, per 1M | !$3.75 | $20 |
| Cost per task? | $0.93 | $0.50 |
| Speed, tok/s | — | 75 |
| Context | 1M | 1.05M |
| Weights | Closed | Closed |
| Released | Sep 2026 | Jul 2026 |
| See images | ✓ | ✓ |
| Read PDFs | ✓ | 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
Gemini 3.8 Flash → GPT-5.6 Sol
Nothing known.
GPT-5.6 Sol → Gemini 3.8 Flash
Context: 1.05M → 1M
Sampling parameters are ignored›
temperature, top_p and top_k are ignored now and will cause HTTP 400 in future generations; remove them from requests.
No minimal thinking level›
The minimal thinking level returns an error; avoid minimal.