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 Sonnet 5Grok 4.3
| Claude Sonnet 5 Anthropic | Grok 4.3 xAI | |
|---|---|---|
| Pricing | ||
| Input $/1M | $2.00 | $1.25 |
| Output $/1M | $10.00 | $2.50 |
| Blended $/1M (3:1) | $4.00 | $1.56 |
| Cache read $/1M | $0.20 | $0.20 |
| Cache write $/1M | $2.50 | — |
| Quality indices | ||
| Intelligence Index | 53.4 | 37.6 |
| Coding Index | 71.5 | 42.2 |
| Math Index | — | — |
| Agentic Index | 46.7 | 24.1 |
| Speed & latency | ||
| Output speed (tok/s) | 86 | 122 |
| Time to first token | 93.73s | 32.04s |
| Time to first answer token | 93.73s | 32.04s |
| Benchmarks (Artificial Analysis) | ||
| GPQA Diamond | 91.1% | 90.1% |
| Humanity's Last Exam | 39.6% | 35.0% |
| SciCode | 53.6% | 47.3% |
| IFBench | — | 81.3% |
| AA-LCR (Long Context Reasoning) | 70.7% | 64.3% |
| Terminal-Bench Hard | — | 37.9% |
| Terminal-Bench 2.1 | 80.5% | 39.7% |
| τ²-Bench (Telecom) | — | 97.7% |
| τ³-Bench Banking | 28.2% | 12.2% |
| Design Arena (Elo) | ||
| Website | 1,314 | 1,221 |
| Web apps | 1,305 | 1,179 |
| Full-stack | 1,292 | 1,070 |
| Mobile apps | — | 1,130 |
| Android native | 1,255 | 1,056 |
| UI components | 1,317 | 1,236 |
| Data viz | 1,281 | 1,222 |
| SVG | 1,239 | 1,128 |
| 3D | 1,321 | 1,182 |
| Game dev | 1,345 | 1,235 |
| Agentic game dev | 1,241 | 1,017 |
| Godot game dev | 1,257 | 1,061 |
| HTML slides | 1,233 | 1,045 |
| PPTX slides | — | 1,071 |
| Python→PPTX slides | 1,252 | 1,072 |
| Agentic slides | — | 1,072 |
| Agentic HTML slides | — | 1,066 |
| Agentic slides (HTML) | — | 1,066 |
| Agentic slides (Python PPTX) | — | 1,068 |
| Code categories | 1,312 | 1,221 |
| ASCII art | 1,230 | 1,184 |
| Specs | ||
| Context window | 1M | 1M |
| Max output tokens | 128K | — |
| Input modalities | text, image, file | text, image, file |
| Tokenizer | Claude | Grok |