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 | 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 | 128k |
| API pricingUSD per 1M tokens · lower wins | ||
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | — | $0.25DeepInfra |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $0.95DeepInfra |
| — | $97.50DeepInfra · fp4 | |
| Benchmarks | ||
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 49.2% | 66% |
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. | 65.9% | 74.8% |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | 79.8% | 93.1% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 71.5% | 80.1% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 84% | 84.8% |
SWE-Bench MultilingualMultilingual coding — Like SWE-Bench, but the coding problems span many programming languages, not just one. Tests how broadly the AI can code. Higher is better. | — | 54.5% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | — | 30% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | — | 15.9% |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 97.3% | — |
DROPReasoning — three-shot token F1 on questions requiring reasoning over paragraphs | 92.2 | — |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | — | 88.4% |
SimpleQAKnowledge — accuracy on short, fact-seeking questions in the original SimpleQA set | 30.1% | — |
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. | 90.8% | — |
| Overview | ||
| Company | DeepSeek | DeepSeek |
| Release date | Jan 20 2025 | Aug 21 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek R1vsClaude Fable 5.2DeepSeek V3.1vsClaude Fable 5.2DeepSeek R1vsGPT-6.2 AstraDeepSeek V3.1vsGPT-6.2 AstraDeepSeek R1vsGemini 4 HydrogenDeepSeek V3.1vsGemini 4 HydrogenDeepSeek R1vsMuse EmberDeepSeek V3.1vsMuse EmberDeepSeek R1vsGrok 4.7DeepSeek V3.1vsGrok 4.7DeepSeek R1vsMistral Large 4DeepSeek V3.1vsMistral Large 4Frequently asked questions
DeepSeek V3.1 leads DeepSeek R1 on 5 of the 5 benchmarks they both report. DeepSeek R1 shipped 213 days before DeepSeek V3.1, so benchmark comparisons should account for the intervening progress.
DeepSeek R1 has 671B parameters, while DeepSeek V3.1 has 671B. Context windows are 128k (DeepSeek R1) vs 128k (DeepSeek V3.1).
On SWE-Bench Verified, DeepSeek V3.1 leads at 66% vs DeepSeek R1 at 49.2%. On LiveCodeBench, DeepSeek V3.1 leads at 74.8% vs DeepSeek R1 at 65.9%. On AIME 2024, DeepSeek V3.1 leads at 93.1% vs DeepSeek R1 at 79.8%. On GPQA Diamond, DeepSeek V3.1 leads at 80.1% vs DeepSeek R1 at 71.5%. On MMLU-Pro, DeepSeek V3.1 leads at 84.8% vs DeepSeek R1 at 84%.