BFL Releases FLUX 3 Video Model, Open Weights Confirmed
cloneofsimo · x · 2026-08-05
Black Forest Labs (BFL) has officially released its first video model, FLUX 3 Video, which has been in development for two years. The team describes it as a strong, versatile model.
Beyond the current release, BFL outlined future updates, including generation from references and a highly anticipated open-weight version. Upcoming variants will also feature action prediction and image generation capabilities.
Related event: Black Forest Labs Launches FLUX 3 Multimodal Model(19 posts)→
More from Multimodal
- Viral MiniMax H3 Demo: AI Noise Mutates into Anime Girl, Smashes GPU — circlenline · 2026-08-05
- InfiniSplat Generates Navigable 3D Scenes from a Single Image in Under 1 Second — victormustar · 2026-08-05
- Minimax H3 Ignores 'Stationary Camera' Prompt, Continues Cutting and Moving — backworld_nograv · 2026-08-05
- RabbitPre Launches RabbitVis: Visual AI Moves from Generation to Editable Delivery — 机器之心 · 2026-08-05
- LTX IC LoRA Test: Video Style Transfer in 5 Minutes on RTX 3090 — Optimal-Spare1305 · 2026-08-05
- ByteDance Integrates Seedance Video Model into Gauth for EdTech Animations — Economy_Cicada8756 · 2026-08-05