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. | — | 6B |
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. | 200k | 8k |
| 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. | 97.3% | 58.5% |
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% | 61.4% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 8% | — |
MATHReasoning — accuracy on the original competition mathematics test set | — | 25.7% |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 97.5% | — |
DROPReasoning — three-shot token F1 on questions requiring reasoning over paragraphs | 77.7 | — |
HellaSwagReasoning — commonsense completion of short scenarios | — | 79.7% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 77.6% | — |
SimpleQAKnowledge — accuracy on short, fact-seeking questions in the original SimpleQA set | 20.2% | — |
MGSMMultilingual math — solving grade-school word problems translated into multiple languages | 93.7% | — |
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. | — | 72.3% |
| Overview | ||
| Company | OpenAI | Z.ai |
| Release date | Apr 16 2025 | Oct 27 2023 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
o4-minivsClaude Fable 5.2ChatGLM3-6BvsClaude Fable 5.2o4-minivsGemini 4 HydrogenChatGLM3-6BvsGemini 4 Hydrogeno4-minivsMuse EmberChatGLM3-6BvsMuse Embero4-minivsGrok 4.7ChatGLM3-6BvsGrok 4.7o4-minivsDeepSeek-V4.1-Flash-1017ChatGLM3-6BvsDeepSeek-V4.1-Flash-1017o4-minivsMistral Large 4ChatGLM3-6BvsMistral Large 4Frequently asked questions
o4-mini leads ChatGLM3-6B on 2 of the 2 benchmarks they both report (HumanEval, MMLU). ChatGLM3-6B shipped 537 days before o4-mini, so benchmark comparisons should account for the intervening progress.
Context windows are 200k (o4-mini) vs 8k (ChatGLM3-6B). o4-mini is closed, while ChatGLM3-6B is open weight.
On HumanEval, o4-mini leads at 97.3% vs ChatGLM3-6B at 58.5%. On MMLU, o4-mini leads at 90% vs ChatGLM3-6B at 61.4%.