Compare AI Models
Side-by-side benchmark scores, prices, speed and context windows — then share the result as an image.
vs
Claude Opus 5.5 wins 6 of 11 head-to-head metrics against GPT-6 Astra.
| Metric | Claude Opus 5.5 | GPT-6 Astra |
|---|---|---|
| AMR Score | 96.8✓ | 95.6 |
| GPQA Diamond | 95.8% | 96%✓ |
| MMLU-Pro | 91.2%✓ | 90.1% |
| SWE-Bench Verified | 81.4%✓ | 78.9% |
| AIME 2025 | 96.7%✓ | 95.8% |
| LiveCodeBench v6 | 84.2%✓ | 82.8% |
| Arena Elo | 1489✓ | 1474 |
| Input price (/1M tokens) | $8 | $5✓ |
| Output price (/1M tokens) | $40 | $30✓ |
| Output speed (tok/s) | 88 | 128✓ |
| Context window | 500K | 1M✓ |
Share this comparison
Copy the card below straight into Slack, X, Discord or Reddit.
AI model rankings FAQ
How do I compare two AI models?
Pick the two models above and the table fills in every tracked metric: five public benchmarks, arena Elo, input and output price, output speed and context window. The stronger value in each row is highlighted, the verdict counts head-to-head wins, and the share button turns the result into an image you can paste anywhere.