Compare DeepSeek V3.2 Exp from DeepSeek and Codestral 2508 from Mistral AI on key metrics including benchmarks, price, context length, and other model features. Access both models and hundreds of others through the OpenRouter API.




DeepSeek V3.2 Exp and Codestral 2508 are available through the OpenRouter API, so switching between them takes a model slug change rather than a new integration.
DeepSeek V3.2 Exp, from DeepSeek, has a 163,840-token context window and is priced at $0.27/M tokens input, $0.41/M tokens output on OpenRouter.
Codestral 2508, from Mistral AI, has a 256,000-token context window and is priced at $0.30/M tokens input, $0.90/M tokens output on OpenRouter.
| Attribute | DeepSeek V3.2 Exp | Codestral 2508 |
|---|---|---|
| Input price | $0.27/M tokens | $0.30/M tokens |
| Output price | $0.41/M tokens | $0.90/M tokens |
| Context window | 163,840 tokens | 256,000 tokens |
| Intelligence Index | 16.6 | Not available |
| Coding Index | Not available | Not available |
| Agentic Index | Not available | Not available |
| Latency (p50) | 2.30 s | 0.28 s |
| Throughput (p50) | 17.0 tok/s | 91.0 tok/s |
Benchmark data updated
DeepSeek V3.2 Exp is cheaper than Codestral 2508 on OpenRouter: DeepSeek V3.2 Exp costs $0.27/M tokens input and $0.41/M tokens output, while Codestral 2508 costs $0.30/M tokens input and $0.90/M tokens output.
Codestral 2508 has the longer context window at 256,000 tokens, compared with 163,840 tokens for DeepSeek V3.2 Exp.
Codestral 2508 is faster on OpenRouter, generating a median 91.0 tokens per second compared with 17.0 tokens per second for DeepSeek V3.2 Exp.