Compare models
Pick up to 4 models for a side-by-side spec sheet. Hover any dotted metric name for what the test measures. Best value in each row is highlighted.
GPT-5.6 SolQwen3.5 397B A17B
| GPT-5.6 Sol OpenAI | Qwen3.5 397B A17B Alibaba (Qwen) | |
|---|---|---|
| Pricing | ||
| Input $/1M | $2.00 | $0.55 |
| Output $/1M | $10.00 | $3.50 |
| Blended $/1M (3:1) | $4.00 | $1.29 |
| Cache read $/1M | $0.20 | $0.225 |
| Cache write $/1M | $2.50 | — |
| Quality indices | ||
| Intelligence Index | 47 | 18.4 |
| Coding Index | 77.4 | 48.2 |
| Math Index | — | — |
| Agentic Index | 50.2 | 8.3 |
| Speed & latency | ||
| Output speed (tok/s) | 65 | 87 |
| Time to first token | 64.46s | 1.59s |
| Time to first answer token | 64.46s | 38.12s |
| Benchmarks (Artificial Analysis) | ||
| GPQA Diamond | 94.1% | 89.3% |
| Humanity's Last Exam | 49.5% | 29.0% |
| SciCode | 57.1% | 44.8% |
| IFBench | 72.7% | 78.8% |
| AA-LCR (Long Context Reasoning) | 84.0% | 77.3% |
| Terminal-Bench Hard | 65.9% | 40.9% |
| Terminal-Bench 2.1 | 88.0% | 51.3% |
| τ²-Bench (Telecom) | 85.1% | 95.6% |
| τ³-Bench Banking | 44.3% | 13.4% |
| Design Arena (Elo) | ||
| Website | — | 1,203 |
| UI components | — | 1,179 |
| Data viz | — | 1,190 |
| SVG | — | 1,163 |
| 3D | — | 1,189 |
| Game dev | — | 1,163 |
| Code categories | — | 1,195 |
| Specs | ||
| Context window | 1.05M | 262K |
| Max output tokens | 128K | 236K |
| Input modalities | file, image, text | text, image, video |
| Tokenizer | GPT | Qwen3 |