Open Systems, Blockchain & Business Daily

OSBBD

AI, money and the software behind them.

DeepSeek-V4.1-Flash

DeepSeek-V4.1-Flash is an open-weight language model from DeepSeek, released on September 10, 2026.

Developer
DeepSeek
Released
September 10, 2026
Total parameters
552B + 196B memory
Active parameters per token
8B–16B
Architecture
MoE
Context window
1M
Input
Text, image
License
MIT
Commercial use
Yes
Official weights, GB
510.3 GB
Hugging Face repository
huggingface.co
Checked on

DeepSeek-V4.1-Flash is a mixture-of-experts model with a 552B-parameter backbone plus 196B parameters of Engram conditional memory. During prefill 8B parameters are active per token, and during decoding 16B. DeepSeek lists a context window of 1M tokens. The model accepts text and image input. The weights are released under the MIT license, which allows commercial use.

The official weights on Hugging Face take about 510 GB. Running the model needs at least that much memory across GPUs and system RAM, plus room for the context cache; quantized versions need less.

Total parameters, billions
748B
Active parameters, billions
8B
Context window, tokens
1M
License type
MIT
License conditions
MIT: keep the copyright and permission notice.
Weights precision
Mixed FP8 + BF16
License text
huggingface.co
Sources
PageURL
Hugging Face model pagehttps://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash
License text on Hugging Facehttps://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash/blob/main/LICENSE
Hugging Face weights indexhttps://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash/raw/main/model.safetensors.index.json
DeepSeek API docshttps://api-docs.deepseek.com/updates

Questions

What license does DeepSeek-V4.1-Flash use?
The weights are released under the MIT license, which allows commercial use.
Can I use DeepSeek-V4.1-Flash commercially?
Yes. The MIT license only asks you to keep the copyright and permission notice.
How much memory does DeepSeek-V4.1-Flash need?
The official DeepSeek-V4.1-Flash weights take about 510 GB (mixed FP8 and BF16). Running the model needs at least that much memory across GPUs and system RAM, plus room for the context cache; quantized versions need less. The memory calculator checks which setups fit.
How many parameters does DeepSeek-V4.1-Flash have?
DeepSeek-V4.1-Flash has a 552B-parameter backbone plus 196B parameters of Engram conditional memory. During prefill 8B parameters are active per token, and during decoding 16B.

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.

OSBBD

What are you looking for?

Search by headline, topic or keyword.