GPQA Diamond
?what it ishighest score on this metric 96.4%
193º / 198 · Epoch AI, latest version we checked
DeepSeek LLM 67B
Benchmarks, price, context and versions
Direct answer: DeepSeek LLM 67B is a model from DeepSeek. 24.6% on GPQA Diamond, per Epoch AI.
✓ Good for
- running on your own infrastructure (open weights)
✕ Not the best for
- technical reasoning on science topics (#193 of 198 on GPQA Diamond)
- calculation and problems with an exact answer (#182 of 186 on OTIS Mock AIME)
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
Key specs
Version history
- deepseek-llm-67b-chat Hugging Face Hub, Hugging Face repo created
- deepseek-llm-67b-base Hugging Face Hub, Hugging Face repo created
- deepseek-llm-7b-chat Hugging Face Hub, Hugging Face repo created
- deepseek-llm-7b-base Hugging Face Hub, Hugging Face repo created
Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch.
FAQ
What is DeepSeek LLM 67B?
DeepSeek LLM 67B is a model from DeepSeek, with open weights, used for text.
How much does DeepSeek LLM 67B cost?
No data in open sources.
What is the context window of DeepSeek LLM 67B?
No data in open sources.
Is DeepSeek LLM 67B open source?
Yes, it has open weights. License: no data.
What is DeepSeek LLM 67B's GPQA Diamond score?
24.6% on GPQA Diamond, per Epoch AI.
Is DeepSeek LLM 67B good for coding?
No data: there is no SWE-bench Verified measurement for it in open sources.
Who made DeepSeek LLM 67B?
It was made by DeepSeek, a lab from China.
provisional data Questions under review: Semrush had no API balance during collection. Answers use source data only.
Models from DeepSeek
Sources
- Epoch AI, 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.