Side-by-side comparison of AI model specs, benchmarks, and pricing.
| Spec | Gemini 2.5 Flash Google | Grok 4 xAI |
|---|---|---|
| Tier | mid | frontier |
| Release Date | 2025-06-17 | 2025-07-09 |
| Context Window | 1049K | 256K |
| Max Output | 66K | 8K |
| Input Price (/1M) | $0.3 | $3 |
| Output Price (/1M) | $2.5 | $15 |
| Arena Elo | 1,410 | 1,410 |
| MMLU | 83.5% | 87.5% |
| GPQA | 65.8% | 87.5% |
| MATH | 82.1% | 91.7% |
| HumanEval | 90.3% | 90% |
| SWE-bench | — | 48% |
| Capabilities | vision, tool-use, code, reasoning, audio | vision, tool-use, code, reasoning |