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 | 4B/30B/235B |
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 | 256k |
| 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% | 74.1% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | 30.6% | 18.2% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 96% | 92.3% |
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% | 81.1% |
IFEvalAgentic tasks — following verifiable instructions in generated responses | — | 87.8% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | — | 84.4% |
| Overview | ||
| Company | DeepSeek | Qwen |
| Release date | Dec 1 2025 | Jul 21 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek-V3.2-SpecialevsClaude Fable 5.2Qwen3-2507vsClaude Fable 5.2DeepSeek-V3.2-SpecialevsGPT-6.2 AstraQwen3-2507vsGPT-6.2 AstraDeepSeek-V3.2-SpecialevsGemini 4 NeonQwen3-2507vsGemini 4 NeonDeepSeek-V3.2-SpecialevsMuse EmberQwen3-2507vsMuse EmberDeepSeek-V3.2-SpecialevsGrok 4.7Qwen3-2507vsGrok 4.7DeepSeek-V3.2-SpecialevsMistral Large 4Qwen3-2507vsMistral Large 4Frequently asked questions
DeepSeek-V3.2-Speciale leads Qwen3-2507 on 4 of the 4 benchmarks they both report (LiveCodeBench, Humanity's Last Exam, AIME 2025, GPQA Diamond). Qwen3-2507 shipped 133 days before DeepSeek-V3.2-Speciale, so benchmark comparisons should account for the intervening progress.
DeepSeek-V3.2-Speciale has 671B parameters, while Qwen3-2507 has 4B/30B/235B. Context windows are 128k (DeepSeek-V3.2-Speciale) vs 256k (Qwen3-2507).
On LiveCodeBench, DeepSeek-V3.2-Speciale leads at 88.7% vs Qwen3-2507 at 74.1%. On Humanity's Last Exam · text only, no tools, DeepSeek-V3.2-Speciale leads at 30.6% vs Qwen3-2507 at 18.2%. On AIME 2025, DeepSeek-V3.2-Speciale leads at 96% vs Qwen3-2507 at 92.3%. On GPQA Diamond, DeepSeek-V3.2-Speciale leads at 85.7% vs Qwen3-2507 at 81.1%.