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 4.8 (batch)GPT-5.6 Sol
| Claude Opus 4.8 (batch) Anthropic | GPT-5.6 Sol OpenAI | |
|---|---|---|
| Pricing | ||
| Input $/1M | $2.50 | $2.00 |
| Output $/1M | $12.50 | $10.00 |
| Blended $/1M (3:1) | $5.00 | $4.00 |
| Cache read $/1M | $0.25 | $0.20 |
| Cache write $/1M | $3.13 | $2.50 |
| Quality indices | ||
| Intelligence Index | 41.8 | 47 |
| Coding Index | 74.3 | 77.4 |
| Math Index | — | — |
| Agentic Index | 41.9 | 50.2 |
| Speed & latency | ||
| Output speed (tok/s) | 0 | 65 |
| Time to first token | 0s | 64.46s |
| Time to first answer token | 0s | 64.46s |
| Benchmarks (Artificial Analysis) | ||
| GPQA Diamond | 92.0% | 94.1% |
| Humanity's Last Exam | 48.7% | 49.5% |
| SciCode | 54.4% | 57.1% |
| IFBench | 62.2% | 72.7% |
| AA-LCR (Long Context Reasoning) | 77.7% | 84.0% |
| Terminal-Bench Hard | 58.3% | 65.9% |
| Terminal-Bench 2.1 | 84.6% | 88.0% |
| τ²-Bench (Telecom) | 94.4% | 85.1% |
| τ³-Bench Banking | 34.2% | 44.3% |
| Design Arena (Elo) | ||
| Website | 1,267 | — |
| Web apps | 1,246 | — |
| Full-stack | 1,256 | — |
| Mobile apps | 1,254 | — |
| Android native | 1,246 | — |
| UI components | 1,270 | — |
| Data viz | 1,254 | — |
| SVG | 1,210 | — |
| 3D | 1,252 | — |
| Game dev | 1,281 | — |
| Agentic game dev | 1,238 | — |
| Godot game dev | 1,236 | — |
| HTML slides | 1,223 | — |
| PPTX slides | 1,306 | — |
| Python→PPTX slides | 1,298 | — |
| Agentic slides | 1,294 | — |
| Agentic HTML slides | 1,227 | — |
| Agentic slides (HTML) | 1,230 | — |
| Agentic slides (Python PPTX) | 1,310 | — |
| Code categories | 1,267 | — |
| ASCII art | 1,295 | — |
| Specs | ||
| Context window | 1M | 1.05M |
| Max output tokens | 128K | 128K |
| Input modalities | text, image, file | file, image, text |
| Tokenizer | Claude | GPT |