Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 236B | 22B |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 128k | 32k |
| Overview | ||
| Company | DeepSeek | Mistral |
| Release date | Sep 5 2024 | May 29 2024 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
DeepSeek V2.5vsClaude Haiku 5.5Codestral 22BvsClaude Haiku 5.5DeepSeek V2.5vsGPT-6.1 SolCodestral 22BvsGPT-6.1 SolDeepSeek V2.5vsGemini 4 ArgonCodestral 22BvsGemini 4 ArgonDeepSeek V2.5vsMuse Spark OpenCodestral 22BvsMuse Spark OpenDeepSeek V2.5vsGrok 4.7Codestral 22BvsGrok 4.7DeepSeek V2.5vsKimi K3Codestral 22BvsKimi K3Frequently asked questions
DeepSeek V2.5 and Codestral 22B don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Codestral 22B shipped 99 days before DeepSeek V2.5, so benchmark comparisons should account for the intervening progress.
DeepSeek V2.5 has 236B parameters, while Codestral 22B has 22B. Context windows are 128k (DeepSeek V2.5) vs 32k (Codestral 22B). DeepSeek V2.5 is open weight, while Codestral 22B is closed.
Direct benchmark comparisons are unavailable — DeepSeek V2.5 and Codestral 22B don't publish scores on any of the same benchmarks.