DeepSeek-V4-Pro-0813
DeepSeek-V4-Pro-0813 is an open-weight language model from DeepSeek, released on August 13, 2026.
- Developer
- DeepSeek
- Released
- August 13, 2026
- Total parameters
- 1.6T
- Active parameters per token
- 49B
- Architecture
- MoE
- Context window
- 1M
- Input
- Text
- License
- MIT
- Commercial use
- Yes
- Hugging Face repository
- huggingface.co
- Checked on
DeepSeek-V4-Pro-0813 is a mixture-of-experts model with 1.6T total parameters, of which 49B are active for each token. DeepSeek lists a context window of 1M tokens. The model takes text input. The weights are released under the MIT license, which allows commercial use.
- Total parameters, billions
- 1,600B
- Active parameters, billions
- 49B
- Context window, tokens
- 1M
- License type
- MIT
- License conditions
- MIT: keep the copyright and permission notice.
- Weights precision
- FP4 + FP8
- License text
- huggingface.co
- Sources
Page URL Hugging Face model card https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813/raw/main/README.md Hugging Face model card https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/raw/main/README.md Hugging Face API https://huggingface.co/api/models/deepseek-ai/DeepSeek-V4-Pro-0813?expand=safetensors DeepSeek API docs https://api-docs.deepseek.com/updates DeepSeek API docs: release notes https://api-docs.deepseek.com/news/news260424
Questions
What license does DeepSeek-V4-Pro-0813 use?
The weights are released under the MIT license, which allows commercial use.
Can I use DeepSeek-V4-Pro-0813 commercially?
Yes. The MIT license only asks you to keep the copyright and permission notice.
How much memory does DeepSeek-V4-Pro-0813 need?
DeepSeek-V4-Pro-0813 has 1.6T total parameters, and all of them must be loaded. As a rough guide, a model takes about 2 GB per billion parameters in BF16, 1 GB in 8-bit and a little over 0.5 GB in 4-bit, before the context cache; the memory calculator checks which setups fit.
How many parameters does DeepSeek-V4-Pro-0813 have?
DeepSeek-V4-Pro-0813 is a mixture-of-experts model with 1.6T total parameters, of which 49B are active for each token. DeepSeek lists a context window of 1M tokens.
Sources
Every figure comes from the developer’s model card, license text and official announcements, and from the model’s Hugging Face repository metadata. Each model page lists its sources and the date they were checked; the register is rechecked every week.