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.
Claude Haiku 4.5gpt-oss-120b (batch)
| Claude Haiku 4.5 Anthropic | gpt-oss-120b (batch) OpenAI | |
|---|---|---|
| Pricing | ||
| Input $/1M | $1.00 | $0.15 |
| Output $/1M | $5.00 | $0.60 |
| Blended $/1M (3:1) | $2.00 | $0.262 |
| Cache read $/1M | $0.10 | — |
| Cache write $/1M | $1.25 | — |
| Quality indices | ||
| Intelligence Index | 16.9 | 11.6 |
| Coding Index | 43.9 | 30.4 |
| Math Index | — | 93.4 |
| Agentic Index | 8 | 3.7 |
| Speed & latency | ||
| Output speed (tok/s) | — | 207 |
| Time to first token | — | 0.52s |
| Time to first answer token | — | 10.18s |
| Benchmarks (Artificial Analysis) | ||
| MMLU-Pro | — | 80.8% |
| GPQA Diamond | — | 78.2% |
| Humanity's Last Exam | — | 19.6% |
| LiveCodeBench | — | 87.8% |
| SciCode | — | 34.0% |
| AIME 2025 | — | 93.4% |
| IFBench | — | 69.0% |
| AA-LCR (Long Context Reasoning) | — | 52.0% |
| Terminal-Bench Hard | — | 23.5% |
| Terminal-Bench 2.1 | — | 26.2% |
| τ²-Bench (Telecom) | — | 65.8% |
| τ³-Bench Banking | — | 12.8% |
| Design Arena (Elo) | ||
| Website | 1,135 | 980 |
| UI components | 1,115 | 942 |
| Data viz | 1,139 | 1,002 |
| SVG | 1,049 | — |
| 3D | 1,101 | 930 |
| Game dev | 1,120 | 1,015 |
| Code categories | 1,130 | 978 |
| ASCII art | 1,159 | — |
| Specs | ||
| Context window | 200K | 131K |
| Max output tokens | 64K | 118K |
| Input modalities | text, image, file | text |
| Tokenizer | Claude | GPT |