Compare models
Comparing 3models. Drop the URL into a doc — it's permalinked.
| Field | openai/o1 | anthropic/claude-opus-4-5 | google/gemini-2.5-pro |
|---|---|---|---|
| Provider | openai | anthropic | |
| Model ID | o1 | claude-opus-4-5 | gemini-2.5-pro |
| Context | 200K | 200K | 1.0M |
| Max output | 100K | 32K | 8K |
| Input / 1M | $15.00 | $15.00 | $1.25 |
| Output / 1M | $60.00 | $75.00 | $10.00 |
| Cached input / 1M | $7.50 | $1.50 | $0.31 |
| Avg cost / 1M | $37.50 | $45.00 | $5.63 |
| Speed | 35 t/s | 65 t/s | 130 t/s |
| Quality index | 85.0 | 80.0 | 80.0 |
| MMLU | 92.3 | 91.4 | 89.2 |
| GPQA | 78.0 | 79.6 | 84.0 |
| HumanEval | 89.0 | 92.7 | 92.6 |
| MATH | 94.8 | 96.5 | 91.6 |
| SWE-bench | 48.9 | 72.5 | 63.8 |
| Arena Elo | — | — | 1380 |
| Tools | ✓ | ✓ | ✓ |
| Vision | ✓ | ✓ | ✓ |
| Thinking | ✓ | ✓ | ✓ |
| Streaming | ✓ | ✓ | ✓ |
| JSON mode | — | — | ✓ |
| Structured output | ✓ | ✓ | ✓ |
| Prompt cache | — | ✓ | — |
Same data, in your terminal: relay models compare o1-pro opus-4.5 gemini-pro