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. | 1.6T | 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. | 1M | 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. | $0.8771StreamLake | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $1.50Relace | — |
| $255Relace · fp4 | — | |
| 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. | 93.5% | 74.1% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 90.1% | 81.1% |
CheatBenchCheating rate — Measures how often AI agents try to cheat on difficult assignments, such as reading hidden answers, copying work or manipulating grading. The overall score gives equal weight to ten categories; the sycophancy category measures how far an agent shifts its beliefs toward a user's stated views. Scores describe each model in its tested agent setup, and task success is measured separately. Lower is better. | 75.1% | — |
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. | 14% | — |
IFEvalAgentic tasks — following verifiable instructions in generated responses | — | 87.8% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 83.4% | — |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | — | 18.2% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | — | 92.3% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | — | 84.4% |
| Overview | ||
| Company | DeepSeek | Qwen |
| Release date | Apr 24 2026 | Jul 21 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek-V4-ProvsClaude Fable 5.2Qwen3-2507vsClaude Fable 5.2DeepSeek-V4-ProvsGPT-6.2 AstraQwen3-2507vsGPT-6.2 AstraDeepSeek-V4-ProvsGemini 4 NeonQwen3-2507vsGemini 4 NeonDeepSeek-V4-ProvsMuse EmberQwen3-2507vsMuse EmberDeepSeek-V4-ProvsGrok 4.7Qwen3-2507vsGrok 4.7DeepSeek-V4-ProvsMistral Large 4Qwen3-2507vsMistral Large 4Frequently asked questions
DeepSeek-V4-Pro leads Qwen3-2507 on 2 of the 2 benchmarks they both report (LiveCodeBench, GPQA Diamond). Qwen3-2507 shipped 277 days before DeepSeek-V4-Pro, so benchmark comparisons should account for the intervening progress.
DeepSeek-V4-Pro has 1.6T parameters, while Qwen3-2507 has 4B/30B/235B. Context windows are 1M (DeepSeek-V4-Pro) vs 256k (Qwen3-2507).
On LiveCodeBench, DeepSeek-V4-Pro leads at 93.5% vs Qwen3-2507 at 74.1%. On GPQA Diamond, DeepSeek-V4-Pro leads at 90.1% vs Qwen3-2507 at 81.1%.