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 | 27B |
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 |
| 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. | $0.259SiliconFlow | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $0.38DeepInfra | — |
| $71.00DeepInfra · 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. | 73.1% | 77.2% |
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. | 70.2% | 71.3% |
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. | 83.3% | 83.9% |
Terminal-Bench 2.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? (Version 2.0 of the test.) Higher is better. | 46.4% | 59.3% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 82.4% | 87.8% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 85% | 86.2% |
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. | 13% | — |
NL2Repo-BenchRepo-level code generation — Tests whether the AI can turn a natural-language requirement into working code across an entire repository, not just produce a single function or patch. Higher is better. | — | 36.2% |
ToolathlonGeneral tool use — Tests how well the AI uses everyday real-world tools and apps to get things done. Higher is better. | 35.2% | — |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 51.4% | — |
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. | — | 24% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | 25.1% | — |
AIME 2026Reasoning — single-answer accuracy on the 2026 invitational mathematics exam without tools | — | 94.1% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 93.1% | — |
CharXiv ReasoningChart reasoning — Can the AI read and reason about complex charts and figures, not just text? Higher is better. | — | 78.4% |
MMMU-ProMultimodal reasoning — A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better. | — | 75.8% |
MMMUMultimodal — Tests the AI on understanding images and text together across many college subjects. Higher is better. | — | 82.9% |
| Overview | ||
| Company | DeepSeek | Qwen |
| Release date | Dec 1 2025 | Apr 22 2026 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek-V3.2vsClaude Fable 5.2Qwen3.6-27BvsClaude Fable 5.2DeepSeek-V3.2vsGPT-6.2 AstraQwen3.6-27BvsGPT-6.2 AstraDeepSeek-V3.2vsGemini 4 NeonQwen3.6-27BvsGemini 4 NeonDeepSeek-V3.2vsMuse EmberQwen3.6-27BvsMuse EmberDeepSeek-V3.2vsGrok 4.7Qwen3.6-27BvsGrok 4.7DeepSeek-V3.2vsMistral Large 4Qwen3.6-27BvsMistral Large 4Frequently asked questions
Qwen3.6-27B leads DeepSeek-V3.2 on 6 of the 6 benchmarks they both report. DeepSeek-V3.2 shipped 142 days before Qwen3.6-27B, so benchmark comparisons should account for the intervening progress.
DeepSeek-V3.2 has 671B parameters, while Qwen3.6-27B has 27B. Context windows are 128k (DeepSeek-V3.2) vs 256k (Qwen3.6-27B).
On SWE-Bench Verified, Qwen3.6-27B leads at 77.2% vs DeepSeek-V3.2 at 73.1%. On SWE-Bench Multilingual, Qwen3.6-27B leads at 71.3% vs DeepSeek-V3.2 at 70.2%. On LiveCodeBench, Qwen3.6-27B leads at 83.9% vs DeepSeek-V3.2 at 83.3%. On Terminal-Bench 2.0, Qwen3.6-27B leads at 59.3% vs DeepSeek-V3.2 at 46.4%. On GPQA Diamond, Qwen3.6-27B leads at 87.8% vs DeepSeek-V3.2 at 82.4%. On MMLU-Pro, Qwen3.6-27B leads at 86.2% vs DeepSeek-V3.2 at 85%.