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 | — |
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% | 79% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 96% | 91.7% |
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% | 87.5% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | — | 43% |
Humanity's Last Exam · no toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better. | — | 25.4% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | 30.6% | — |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | — | 38.6% |
| Overview | ||
| Company | DeepSeek | SpaceXSI |
| Release date | Dec 1 2025 | Jul 9 2025 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
DeepSeek-V3.2-SpecialevsClaude Fable 5.2Grok 4vsClaude Fable 5.2DeepSeek-V3.2-SpecialevsGPT-6.2 AstraGrok 4vsGPT-6.2 AstraDeepSeek-V3.2-SpecialevsGemini 4 HydrogenGrok 4vsGemini 4 HydrogenDeepSeek-V3.2-SpecialevsMuse EmberGrok 4vsMuse EmberDeepSeek-V3.2-SpecialevsMistral Large 4Grok 4vsMistral Large 4DeepSeek-V3.2-SpecialevsKimi K3.1 ProGrok 4vsKimi K3.1 ProFrequently asked questions
DeepSeek-V3.2-Speciale leads Grok 4 on 2 of the 3 benchmarks they both report (LiveCodeBench, AIME 2025, GPQA Diamond). Grok 4 shipped 145 days before DeepSeek-V3.2-Speciale, so benchmark comparisons should account for the intervening progress.
Context windows are 128k (DeepSeek-V3.2-Speciale) vs 256k (Grok 4). DeepSeek-V3.2-Speciale is open weight, while Grok 4 is closed.
On LiveCodeBench, DeepSeek-V3.2-Speciale leads at 88.7% vs Grok 4 at 79%. On AIME 2025, DeepSeek-V3.2-Speciale leads at 96% vs Grok 4 at 91.7%. On GPQA Diamond, Grok 4 leads at 87.5% vs DeepSeek-V3.2-Speciale at 85.7%.