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. | — | 2.4T |
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 | 262k |
| Benchmarks | ||
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 81.3% | 92.55% |
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. | 99.3% | — |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | — | 76.03% |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 98.2% | — |
DROPReasoning — three-shot token F1 on questions requiring reasoning over paragraphs | 78.1 | — |
IFBenchInstruction following — Tests whether the AI can follow detailed instructions and satisfy multiple constraints at once. Higher is better. | — | 79.93% |
SimpleQAKnowledge — accuracy on short, fact-seeking questions in the original SimpleQA set | 19.3% | — |
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. | 90.3% | — |
MGSMMultilingual math — solving grade-school word problems translated into multiple languages | 93.5% | — |
| Overview | ||
| Company | OpenAI | Qwen |
| Release date | Apr 16 2025 | Aug 12 2026 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
o4-mini-highvsClaude Fable 5.2Qwen3.8-2.4T-A95BvsClaude Fable 5.2o4-mini-highvsGemini 4 NeonQwen3.8-2.4T-A95BvsGemini 4 Neono4-mini-highvsMuse EmberQwen3.8-2.4T-A95BvsMuse Embero4-mini-highvsGrok 4.7Qwen3.8-2.4T-A95BvsGrok 4.7o4-mini-highvsDeepSeek-V4.1-Flash-1017Qwen3.8-2.4T-A95BvsDeepSeek-V4.1-Flash-1017o4-mini-highvsMistral Large 4Qwen3.8-2.4T-A95BvsMistral Large 4Frequently asked questions
Qwen3.8-2.4T-A95B leads o4-mini-high on 1 of the 1 benchmark they both report (GPQA Diamond). o4-mini-high shipped 483 days before Qwen3.8-2.4T-A95B, so benchmark comparisons should account for the intervening progress.
Context windows are 200k (o4-mini-high) vs 262k (Qwen3.8-2.4T-A95B). o4-mini-high is closed, while Qwen3.8-2.4T-A95B is open weight.
On GPQA Diamond, Qwen3.8-2.4T-A95B leads at 92.55% vs o4-mini-high at 81.3%.