FLUX.3 Video
Black Forest LabsSpecialized
Vision
About this model
FLUX.3 Video is a video generation model from Black Forest Labs. It supports text-to-video, image-guided generation with opening and closing keyframes, and video continuation workflows, making it suited for controlled...
Performance Tier
Specialized
FLUX.3 Video is a specialized model from Black Forest Labs : built for a specific domain.
Domain-specific model. Optimized for a particular task such as code generation, image creation, or web search.
Pricing
This model is included in Elosia plansPremium
Highest cost level. A long conversation can quickly consume your monthly cap.
Capabilities
Context Length—
Max Output Tokens—
Inputtext, image
Outputvideo
Release DateAugust 4, 2026
Recommended Use Cases
Creative Writing
Strengths
- Two rendering tiers picked per generation, 720p or 1080p, set on the request rather than fixed on the model
- Video continuation as a mode of its own, extending an existing clip at 720p or 1080p instead of regenerating it from scratch
- Twenty seconds with native synchronized audio and lip sync documented in thirteen languages, with several shots in a single generation and opening / closing keyframe control
- Single multimodal backbone: one set of weights trained jointly on image, video and audio, not a chain of separate models
Limitations
- No independent validation and no technical report: neither methodology, sample size nor number of raters is published, and the quality claims rest entirely on comparisons run by the vendor
- No seed exposed, so no deterministic reproducibility from one generation to the next. The 1080p output comes from upscaling rather than a native render, and clips are capped at twenty seconds per generation
Frequently asked questions
Resources
This model may use your data for training
