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 Opus 4.7GPT-4.1 Mini
| Claude Opus 4.7 Anthropic | GPT-4.1 Mini OpenAI | |
|---|---|---|
| Pricing | ||
| Input $/1M | $5.00 | $0.40 |
| Output $/1M | $25.00 | $1.60 |
| Blended $/1M (3:1) | $10.00 | $0.70 |
| Cache read $/1M | $0.50 | $0.10 |
| Cache write $/1M | $6.25 | — |
| Quality indices | ||
| Intelligence Index | 53.5 | 14.8 |
| Coding Index | 73.6 | 20.2 |
| Math Index | — | 46.3 |
| Agentic Index | 44.4 | 1.7 |
| Speed & latency | ||
| Output speed (tok/s) | 61 | 83 |
| Time to first token | 16.85s | 0.55s |
| Time to first answer token | 16.85s | 0.55s |
| Benchmarks (Artificial Analysis) | ||
| MMLU-Pro | — | 78.1% |
| GPQA Diamond | 91.4% | 66.4% |
| Humanity's Last Exam | 39.6% | 4.6% |
| LiveCodeBench | — | 48.3% |
| SciCode | 54.5% | 40.4% |
| MATH-500 | — | 92.5% |
| AIME | — | 43.0% |
| AIME 2025 | — | 46.3% |
| IFBench | 58.6% | 38.3% |
| AA-LCR (Long Context Reasoning) | 70.3% | 42.3% |
| Terminal-Bench Hard | 51.5% | 7.6% |
| Terminal-Bench 2.1 | 83.1% | 10.1% |
| τ²-Bench (Telecom) | 88.6% | 52.9% |
| τ³-Bench Banking | 28.9% | 5.2% |
| Design Arena (Elo) | ||
| Website | 1,322 | 1,025 |
| Web apps | 1,311 | — |
| Full-stack | 1,503 | — |
| Android native | 1,266 | — |
| UI components | 1,345 | 1,004 |
| Data viz | 1,315 | 1,070 |
| SVG | 1,272 | — |
| 3D | 1,304 | 895 |
| Game dev | 1,336 | 1,129 |
| Agentic game dev | 1,259 | — |
| Godot game dev | 1,266 | — |
| HTML slides | 1,240 | — |
| PPTX slides | 1,333 | — |
| Python→PPTX slides | 1,336 | — |
| Agentic slides | 1,334 | — |
| Agentic HTML slides | 1,243 | — |
| Agentic slides (HTML) | 1,242 | — |
| Agentic slides (Python PPTX) | 1,331 | — |
| Code categories | 1,321 | 1,027 |
| ASCII art | 1,326 | — |
| Specs | ||
| Context window | 1M | 1.05M |
| Max output tokens | 128K | 33K |
| Input modalities | text, image, file | image, text, file |
| Tokenizer | Claude | GPT |