Side-by-side comparison of AI model specs, benchmarks, and pricing.
| Spec | Gemini 3.1 Pro Google | GPT-4.1 OpenAI |
|---|---|---|
| Tier | frontier | mid |
| Release Date | 2026-02-19 | 2025-04-14 |
| Context Window | 1049K | 1048K |
| Max Output | 66K | 33K |
| Input Price (/1M) | $2 | $2 |
| Output Price (/1M) | $12 | $8 |
| Arena Elo | 1,500 | 1,413 |
| MMLU | 92.6% | 86% |
| GPQA | 94.3% | 66.3% |
| MATH | 96.8% | 73.8% |
| HumanEval | 94.6% | 92.4% |
| SWE-bench | 80.6% | 54.6% |
| AIME | 91.2% | 52% |
| SimpleQA | 79.6% | 42.8% |
| Capabilities | vision, tool-use, code, reasoning, agentic, audio | vision, tool-use, code |