GPQA Diamond
?what it ishighest score on this metric 96.4%
111º / 198 · Epoch AI, latest version we checked
DeepSeek V3
Benchmarks, price, context and versions
Also searched as
DeepSeek V3, DeepSeek V3 vs GPT, deepseek v3.2, deepseek v3 technical report, deepseek v3 0324
Direct answer: DeepSeek V3 is a model from DeepSeek. 67.6% on GPQA Diamond, per Epoch AI. On OpenRouter, US$ 0.29 per 1 million input tokens and US$ 1.14 per 1 million output tokens.
✓ Good for
- running on your own infrastructure (open weights)
✕ Not the best for
- technical reasoning on science topics (#111 of 198 on GPQA Diamond)
- calculation and problems with an exact answer (#126 of 186 on OTIS Mock AIME)
- technical reasoning as the model appears through the API (#110 of 135 on GPQA Diamond (OpenRouter))
- agents that follow rules and use tools (#120 of 131 on τ-bench Verified (airline))
This block follows a rule of this section, using source data: if the model is in the top 25% of the metric, the item goes under "Good for"; if it is in the bottom half, it goes under "Not the best for". The middle band appears in neither. Price and context use OpenRouter; scores use Epoch AI or OpenRouter Benchmarks, in the latest version we checked.
Benchmarks
OTIS Mock AIME
?highest score on this metric 100%
126º / 186 · Epoch AI, latest version we checked
GPQA Diamond (OpenRouter)
?highest score on this metric 96.5%
110º / 135 · OpenRouter Benchmarks, latest version we checked
τ-bench Verified (airline)
?highest score on this metric 82%
120º / 131 · OpenRouter Benchmarks, latest version we checked
Key specs
Version history
- DeepSeek: DeepSeek V3.2 OpenRouter, listed on OpenRouter
- DeepSeek-V3.2 Hugging Face Hub, Hugging Face repo created
- DeepSeek-V3.2-Speciale Hugging Face Hub, Hugging Face repo created
- DeepSeek: DeepSeek V3.2 Exp OpenRouter, listed on OpenRouter
- DeepSeek-V3.2-Exp Hugging Face Hub, Hugging Face repo created
- DeepSeek-V3.2-Exp-Base Hugging Face Hub, Hugging Face repo created
- DeepSeek: DeepSeek V3.1 Terminus OpenRouter, listed on OpenRouter
- DeepSeek-V3.1-Terminus Hugging Face Hub, Hugging Face repo created
- DeepSeek: DeepSeek V3.1 OpenRouter, listed on OpenRouter
- DeepSeek-V3.1 Hugging Face Hub, Hugging Face repo created
- DeepSeek-V3.1-Base Hugging Face Hub, Hugging Face repo created
- DeepSeek: DeepSeek V3 0324 OpenRouter, listed on OpenRouter
- DeepSeek-V3-0324 Hugging Face Hub, Hugging Face repo created
- DeepSeek: DeepSeek V3 OpenRouter, listed on OpenRouter
Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch.
FAQ
How much does DeepSeek V3 cost?
On OpenRouter, US$ 0.29 per 1 million input tokens and US$ 1.14 per 1 million output tokens.
What is DeepSeek V3?
DeepSeek V3 is a model from DeepSeek, with open weights, used for text, agents.
What is the context window of DeepSeek V3?
163,840 tokens, per OpenRouter.
Is DeepSeek V3 open source?
Yes, it has open weights. License: no data.
What is DeepSeek V3's GPQA Diamond score?
67.6% on GPQA Diamond, per Epoch AI.
Is DeepSeek V3 good for coding?
No data: there is no SWE-bench Verified measurement for it in open sources.
Who made DeepSeek V3?
It was made by DeepSeek, a lab from China.
provisional data Questions ordered by what shows up in public Google search suggestions (no volume). Answers use source data only.
Compare with
Models from DeepSeek
Sources
- Epoch AI, CC BY 4.0, latest version we checked.
- OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.
- OpenRouter Benchmarks, CC BY 4.0, latest version we checked.
Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.