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. | — | 9B |
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. | 40k | 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. | 59.36% | 28.5% |
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. | — | 82.3% |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | 73.59% | — |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 64.95% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 70.83% | — |
| Overview | ||
| Company | Mistral | Z.ai |
| Release date | Jun 10 2025 | Jul 5 2024 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
Magistral MediumvsClaude Fable 5.2CodeGeeX4-ALL-9BvsClaude Fable 5.2Magistral MediumvsGPT-6.2 AstraCodeGeeX4-ALL-9BvsGPT-6.2 AstraMagistral MediumvsGemini 4 HydrogenCodeGeeX4-ALL-9BvsGemini 4 HydrogenMagistral MediumvsMuse EmberCodeGeeX4-ALL-9BvsMuse EmberMagistral MediumvsGrok 4.7CodeGeeX4-ALL-9BvsGrok 4.7Magistral MediumvsDeepSeek-V4.1-Flash-1017CodeGeeX4-ALL-9BvsDeepSeek-V4.1-Flash-1017Frequently asked questions
Magistral Medium leads CodeGeeX4-ALL-9B on 1 of the 1 benchmark they both report (LiveCodeBench). CodeGeeX4-ALL-9B shipped 340 days before Magistral Medium, so benchmark comparisons should account for the intervening progress.
Context windows are 40k (Magistral Medium) vs 128k (CodeGeeX4-ALL-9B). Magistral Medium is closed, while CodeGeeX4-ALL-9B is open weight.
On LiveCodeBench, Magistral Medium leads at 59.36% vs CodeGeeX4-ALL-9B at 28.5%.