Compare models
Comparing 2models. Drop the URL into a doc — it's permalinked.
| Field | openai/o3-mini | google/gemini-2.5-pro |
|---|---|---|
| Provider | openai | |
| Model ID | o3-mini | gemini-2.5-pro |
| Context | 200K | 1.0M |
| Max output | 100K | 8K |
| Input / 1M | $1.10 | $1.25 |
| Output / 1M | $4.40 | $10.00 |
| Cached input / 1M | $0.55 | $0.31 |
| Avg cost / 1M | $2.75 | $5.63 |
| Speed | 95 t/s | 130 t/s |
| Quality index | 78.0 | 80.0 |
| MMLU | 86.5 | 89.2 |
| GPQA | 79.7 | 84.0 |
| HumanEval | 87.8 | 92.6 |
| MATH | 97.9 | 91.6 |
| SWE-bench | — | 63.8 |
| Arena Elo | — | 1380 |
| Tools | ✓ | ✓ |
| Vision | — | ✓ |
| Thinking | ✓ | ✓ |
| Streaming | ✓ | ✓ |
| JSON mode | — | ✓ |
| Structured output | ✓ | ✓ |
| Prompt cache | — | — |
Same data, in your terminal: relay models compare openai/o3-mini gemini-pro