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.
MiniMax M3 (batch)GPT-5.6 Sol
| MiniMax M3 (batch) MiniMax | GPT-5.6 Sol OpenAI | |
|---|---|---|
| Pricing | ||
| Input $/1M | $0.30 | $2.00 |
| Output $/1M | $1.20 | $10.00 |
| Blended $/1M (3:1) | $0.525 | $4.00 |
| Cache read $/1M | $0.06 | $0.20 |
| Cache write $/1M | — | $2.50 |
| Quality indices | ||
| Intelligence Index | 29.2 | 47 |
| Coding Index | 58.6 | 77.4 |
| Math Index | — | — |
| Agentic Index | 29.5 | 50.2 |
| Speed & latency | ||
| Output speed (tok/s) | 107 | 65 |
| Time to first token | 1.01s | 64.46s |
| Time to first answer token | 19.78s | 64.46s |
| Benchmarks (Artificial Analysis) | ||
| GPQA Diamond | 92.9% | 94.1% |
| Humanity's Last Exam | 39.0% | 49.5% |
| SciCode | 47.1% | 57.1% |
| IFBench | 82.9% | 72.7% |
| AA-LCR (Long Context Reasoning) | 83.0% | 84.0% |
| Terminal-Bench Hard | 42.4% | 65.9% |
| Terminal-Bench 2.1 | 65.2% | 88.0% |
| τ²-Bench (Telecom) | 88.9% | 85.1% |
| τ³-Bench Banking | 15.3% | 44.3% |
| Design Arena (Elo) | ||
| Website | 1,272 | — |
| Web apps | 1,215 | — |
| Full-stack | 1,204 | — |
| Mobile apps | 1,195 | — |
| Android native | 1,167 | — |
| UI components | 1,261 | — |
| Data viz | 1,250 | — |
| SVG | 1,191 | — |
| 3D | 1,238 | — |
| Game dev | 1,239 | — |
| Agentic game dev | 1,158 | — |
| HTML slides | 1,183 | — |
| Python→PPTX slides | 1,207 | — |
| Code categories | 1,265 | — |
| ASCII art | 1,183 | — |
| Specs | ||
| Context window | 524K | 1.05M |
| Max output tokens | 472K | 128K |
| Input modalities | text, image, video | file, image, text |
| Tokenizer | Other | GPT |