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 | 2.4T |
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 | 262k |
| 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. | $0.27SiliconFlow | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $1.00SiliconFlow | — |
| $104SiliconFlow · fp8 | — | |
| Benchmarks | ||
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 80.7% | 92.55% |
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. | 68.4% | — |
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. | 57.8% | — |
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. | 74.9% | — |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | — | 76.03% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 38.5% | — |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 88.4% | — |
IFBenchInstruction following — Tests whether the AI can follow detailed instructions and satisfy multiple constraints at once. Higher is better. | — | 79.93% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 85% | — |
| Overview | ||
| Company | DeepSeek | Qwen |
| Release date | Sep 22 2025 | Aug 12 2026 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek V3.1 TerminusvsClaude Fable 5.2Qwen3.8-2.4T-A95BvsClaude Fable 5.2DeepSeek V3.1 TerminusvsGPT-6.2 AstraQwen3.8-2.4T-A95BvsGPT-6.2 AstraDeepSeek V3.1 TerminusvsGemini 4 NeonQwen3.8-2.4T-A95BvsGemini 4 NeonDeepSeek V3.1 TerminusvsMuse EmberQwen3.8-2.4T-A95BvsMuse EmberDeepSeek V3.1 TerminusvsGrok 4.7Qwen3.8-2.4T-A95BvsGrok 4.7DeepSeek V3.1 TerminusvsMistral Large 4Qwen3.8-2.4T-A95BvsMistral Large 4Frequently asked questions
Qwen3.8-2.4T-A95B leads DeepSeek V3.1 Terminus on 1 of the 1 benchmark they both report (GPQA Diamond). DeepSeek V3.1 Terminus shipped 324 days before Qwen3.8-2.4T-A95B, so benchmark comparisons should account for the intervening progress.
DeepSeek V3.1 Terminus has 671B parameters, while Qwen3.8-2.4T-A95B has 2.4T. Context windows are 128k (DeepSeek V3.1 Terminus) vs 262k (Qwen3.8-2.4T-A95B).
On GPQA Diamond, Qwen3.8-2.4T-A95B leads at 92.55% vs DeepSeek V3.1 Terminus at 80.7%.