Codestral 25.08 vs GPT-5.4 nano
Price and specs side by side · USD per 1M tokens · verified 2026-10-11
// side by side
| metric | Codestral 25.08 | GPT-5.4 nano |
|---|
| input / 1M | $0.30 | $0.20 |
| output / 1M | $0.90 | $1.25 |
| blended (3:1) | $0.45 | $0.4625 |
| cache read / 1M | $0.03 | $0.02 |
| context | 256K | 400K |
| max output | 8.2K | 128K |
| input types | text | text · image |
| reasoning | no | yes |
| tool calling | yes | yes |
| structured output | yes | yes |
| open weights | no | no |
| released | 2025-07-30 | 2026-03-17 |
| knowledge cutoff | n/a | 2025-08-31 |
| lab | Mistral | OpenAI |
// monthly cost scenarios
| workload | tokens in / out | Codestral 25.08 | GPT-5.4 nano |
|---|
| Chatbot | 10M / 2M | $4.80 | $4.50 |
| RAG app | 50M / 5M | $19.50 | $16.25 |
| Coding agent | 20M / 10M | $15.00 | $16.50 |
// verdict
- Codestral 25.08 is 2.7% cheaper than GPT-5.4 nano on a blended 3:1 input/output basis.
- GPT-5.4 nano has the larger context window (400K vs 256K tokens).
- Only GPT-5.4 nano supports built-in reasoning.
- Only GPT-5.4 nano accepts image input.