DeepSeek-V4.1-Flash appears on Hugging Face: MIT-licensed multimodal model with FP8 weights
AIFlow_ML · x · 2026-09-10
A new repo, DeepSeek-V4.1-Flash, has appeared under the official DeepSeek org on Hugging Face (codename BALENA): pipeline image-text-to-text with text generation, MIT license, 8-bit/FP8 weights, transformers-compatible and endpoints-ready, already at 184 likes within hours of creation. No official details yet — treat as an unverified early sighting until DeepSeek publishes specs.
Related event: DeepSeek Releases V4.1-Flash: 552B MoE with Native Vision, MIT-Licensed(23 posts)→
More from Models
- DeepSeek's new open model beats GLM 5.3 and Kimi K3 at 4-10x lower price — deedydas · 2026-09-10
- DeepSeek eyed for flash-model focus as minor update beats prior Pro model — zephyr_z9 · 2026-09-10
- Screenshot surfaces rare admission of 72-hour KV cache limits in V4-era architecture — zephyr_z9 · 2026-09-10
- User says OpenAI shut down his protein design project, pivoting to open-weight models — Terminator857 · 2026-09-10
- Devs call for standardized "model performance across harnesses" evals — zainhas · 2026-09-10
- DeepSeek V4.1 Flash's reasoning_effort scales output quality and tokens ~linearly in tests — zainhas · 2026-09-10