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. | 671B | 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 |
| Benchmarks | ||
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | 65.9% | 31.5% |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 49.2% | — |
HumanEvalFunction synthesis — 164 small Python problems: the AI is given a function's description and has to write the working function. This was the coding benchmark of the GPT-3.5 and GPT-4 era, before the field moved to fixing real bugs in real repositories. Higher is better. | — | 81.1% |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 97.3% | — |
DROPReasoning — three-shot token F1 on questions requiring reasoning over paragraphs | 92.2 | — |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | 79.8% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 71.5% | — |
SimpleQAKnowledge — accuracy on short, fact-seeking questions in the original SimpleQA set | 30.1% | — |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 84% | — |
MMLUGeneral knowledge — A 57-subject multiple-choice exam — history, law, medicine, maths — that was the standard measure of how much a model knows from 2020 until roughly 2024, when frontier scores crowded into the high 80s and labs moved on to harder tests. The scores here were published years apart under different testing setups, so read them as a historical record rather than a like-for-like ranking. Higher is better. | 90.8% | — |
| Overview | ||
| Company | DeepSeek | Mistral |
| Release date | Jan 20 2025 | May 29 2024 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek R1vsClaude Fable 5.2Codestral 22BvsClaude Fable 5.2DeepSeek R1vsGPT-6.2 AstraCodestral 22BvsGPT-6.2 AstraDeepSeek R1vsGemini 4 HydrogenCodestral 22BvsGemini 4 HydrogenDeepSeek R1vsMuse EmberCodestral 22BvsMuse EmberDeepSeek R1vsGrok 4.7Codestral 22BvsGrok 4.7DeepSeek R1vsKimi K3.1 ProCodestral 22BvsKimi K3.1 ProFrequently asked questions
DeepSeek R1 leads Codestral 22B on 1 of the 1 benchmark they both report (LiveCodeBench). Codestral 22B shipped 236 days before DeepSeek R1, so benchmark comparisons should account for the intervening progress.
DeepSeek R1 has 671B parameters, while Codestral 22B has 22B. Context windows are 128k (DeepSeek R1) vs 32k (Codestral 22B).
On LiveCodeBench, DeepSeek R1 leads at 65.9% vs Codestral 22B at 31.5%.