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. | 32k | 128k |
| Benchmarks | ||
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. | 45.1% | 82.3% |
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. | — | 28.5% |
ARC-ChallengeReasoning — multiple-choice science questions from the AI2 Reasoning Challenge; unrelated to ARC-AGI | 94.2% | — |
HellaSwagReasoning — commonsense completion of short scenarios | 89.2% | — |
WinoGrandeReasoning — resolving ambiguous pronouns in context | 86.7% | — |
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. | 81.2% | — |
GSM8KGrade-school math — Grade-school maths word problems that take a few steps of arithmetic to work through. It separated the models of 2022 and 2023 sharply, then saturated. One caveat on the historical numbers: OpenAI included part of the GSM8K training set in GPT-4's pre-training mix, so GPT-4's score is not a clean few-shot result. Higher is better. | 81% | — |
| Overview | ||
| Company | Mistral | Z.ai |
| Release date | Feb 26 2024 | Jul 5 2024 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
Mistral LargevsClaude Fable 5.2CodeGeeX4-ALL-9BvsClaude Fable 5.2Mistral LargevsGPT-6.2 AstraCodeGeeX4-ALL-9BvsGPT-6.2 AstraMistral LargevsGemini 4 HydrogenCodeGeeX4-ALL-9BvsGemini 4 HydrogenMistral LargevsMuse EmberCodeGeeX4-ALL-9BvsMuse EmberMistral LargevsGrok 4.7CodeGeeX4-ALL-9BvsGrok 4.7Mistral LargevsDeepSeek-V4.1-Flash-1017CodeGeeX4-ALL-9BvsDeepSeek-V4.1-Flash-1017Frequently asked questions
CodeGeeX4-ALL-9B leads Mistral Large on 1 of the 1 benchmark they both report (HumanEval). Mistral Large shipped 130 days before CodeGeeX4-ALL-9B, so benchmark comparisons should account for the intervening progress.
Context windows are 32k (Mistral Large) vs 128k (CodeGeeX4-ALL-9B). Mistral Large is closed, while CodeGeeX4-ALL-9B is open weight.
On HumanEval, CodeGeeX4-ALL-9B leads at 82.3% vs Mistral Large at 45.1%.