model · DeepSeek · open · source listing 2026-09-10
DeepSeek V4.1 Flash
DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on...
trending on Hugging Face3.2k likes430k downloads (30 days)
Sign in to save- maker
- DeepSeek
- openness
- open
- license
- mit
- params b
- 763.2B
- context
- 1M tokens
- price in
- $0.15 / M tokens
- price out
- $0.6 / M tokens
- input modalities
- text, image
- output modalities
- text
- run
- api, local
deepseek · text · vision