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 | — |
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 | 128k |
| 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. | 88.7% | 75% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 96% | 83.48% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 85.7% | 76.26% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | 30.6% | — |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | — | 91.82% |
| Overview | ||
| Company | DeepSeek | Mistral |
| Release date | Dec 1 2025 | Sep 17 2025 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
DeepSeek-V3.2-SpecialevsClaude Fable 5.2Magistral Medium 1.2vsClaude Fable 5.2DeepSeek-V3.2-SpecialevsGPT-6.2 AstraMagistral Medium 1.2vsGPT-6.2 AstraDeepSeek-V3.2-SpecialevsGemini 4 HydrogenMagistral Medium 1.2vsGemini 4 HydrogenDeepSeek-V3.2-SpecialevsMuse EmberMagistral Medium 1.2vsMuse EmberDeepSeek-V3.2-SpecialevsGrok 4.7Magistral Medium 1.2vsGrok 4.7DeepSeek-V3.2-SpecialevsKimi K3.1 ProMagistral Medium 1.2vsKimi K3.1 ProFrequently asked questions
DeepSeek-V3.2-Speciale leads Magistral Medium 1.2 on 3 of the 3 benchmarks they both report (LiveCodeBench, AIME 2025, GPQA Diamond). Magistral Medium 1.2 shipped 75 days before DeepSeek-V3.2-Speciale, so benchmark comparisons should account for the intervening progress.
Context windows are 128k (DeepSeek-V3.2-Speciale) vs 128k (Magistral Medium 1.2). DeepSeek-V3.2-Speciale is open weight, while Magistral Medium 1.2 is closed.
On LiveCodeBench, DeepSeek-V3.2-Speciale leads at 88.7% vs Magistral Medium 1.2 at 75%. On AIME 2025, DeepSeek-V3.2-Speciale leads at 96% vs Magistral Medium 1.2 at 83.48%. On GPQA Diamond, DeepSeek-V3.2-Speciale leads at 85.7% vs Magistral Medium 1.2 at 76.26%.