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 SolQwen3.8 2.4T A95B
| GPT-5.6 Sol OpenAI | Qwen3.8 2.4T A95B Alibaba (Qwen) | |
|---|---|---|
| Pricing | ||
| Input $/1M | $2.00 | $2.00 |
| Output $/1M | $10.00 | $6.00 |
| Blended $/1M (3:1) | $4.00 | $3.00 |
| Cache read $/1M | $0.20 | $0.25 |
| Cache write $/1M | $2.50 | — |
| Quality indices | ||
| Intelligence Index | 47 | 39.9 |
| Coding Index | 77.4 | 71.9 |
| Math Index | — | — |
| Agentic Index | 50.2 | 50.1 |
| Speed & latency | ||
| Output speed (tok/s) | 65 | 39 |
| Time to first token | 64.46s | 1.88s |
| Time to first answer token | 64.46s | 53.08s |
| Benchmarks (Artificial Analysis) | ||
| GPQA Diamond | 94.1% | 93.5% |
| Humanity's Last Exam | 49.5% | 42.4% |
| SciCode | 57.1% | 54.1% |
| IFBench | 72.7% | — |
| AA-LCR (Long Context Reasoning) | 84.0% | 80.3% |
| Terminal-Bench Hard | 65.9% | — |
| Terminal-Bench 2.1 | 88.0% | 82.0% |
| τ²-Bench (Telecom) | 85.1% | — |
| τ³-Bench Banking | 44.3% | 49.1% |
| Specs | ||
| Context window | 1.05M | 1.05M |
| Max output tokens | 128K | 131K |
| Input modalities | file, image, text | text |
| Tokenizer | GPT | Qwen |