Claude Sonnet 4.6 vs GPT-5.2
Lado a lado. Cada métrica. Cada benchmark.
| Tipo | Claude Sonnet 4.6 | GPT-5.2 |
|---|---|---|
| Provider |  Anthropic |  OpenAI |
| pontuação média | 47.6 | 54.0 |
| Preço de entrada | $3.00 | $1.75 |
| Preço de saída | $15.00 | $14.00 |
| Janela de contexto | 1.0M tokens (~500 books) | 400K tokens (~200 books) |
| Lançado em | 2026-02-17 | 2025-12-10 |
| Código aberto | Proprietary | Proprietary |
Pontuações de benchmark
13 benchmarks · Claude Sonnet 4.6: 5, GPT-5.2: 8
| Benchmark | Categoria | Claude Sonnet 4.6 | GPT-5.2 |
|---|---|---|---|
| ARC-AGI | reasoning | 86.5 | 86.2 |
| ARC-AGI-2 | reasoning | 60.4 | 52.9 |
| Chatbot Arena Elo — Coding | arena | 1521.0 | 1403.1 |
| Chatbot Arena Elo — Overall | arena | 1462.2 | 1439.5 |
| Chess Puzzles | knowledge | 13.0 | 49.0 |
| FrontierMath-2025-02-28-Private | math | 32.4 | 40.7 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 8.3 | 18.8 |
| GPQA diamond | knowledge | 83.2 | 88.5 |
| OTIS Mock AIME 2024-2025 | math | 85.8 | 96.1 |
| PostTrainBench | knowledge | 16.4 | 21.4 |
| SimpleQA Verified | knowledge | 29.0 | 38.9 |
| SWE-Bench verified | coding | 75.2 | 73.8 |
| WeirdML | coding | 66.1 | 72.2 |