Side-by-side comparison of Codestral (Mistral AI) and GPT-4o (OpenAI) — benchmarks, pricing, and capabilities.
| Codestral Mistral AI | GPT-4o OpenAI | |
|---|---|---|
| Category | Code Models | LLMs |
| Specifications | ||
| Context Window | 32K | 128K |
| Pricing (per 1M tokens) | ||
| Input Cost | $0.20 | $2.50 |
| Output Cost | $0.60 | $10.00 |
| Performance | ||
| Overall Score | 86.2 | 92.5 |
| ARC-Challenge | — | 88.7 |
| BigBench Hard | — | 87.2 |
| Chatbot Arena ELO | — | 1164.0 |
| DROP | — | 88.1 |
| GSM8K | 84.0 | 72.0 |
| HumanEval | 92.5 | 92.0 |
| MATH | 60.5 | 76.6 |
| MBPP | 88.0 | — |
| MMLU | 79.0 | 88.7 |
| TruthfulQA | — | 72.2 |
| WinoGrande | — | 85.2 |
| Community | ||
| User Rating | ★ 4.6 | ★ 4.7 |
| Reviews | 289 | 1240 |