
FLUX.1 [dev] Variants: 5
The open-weights FLUX workhorse, sharper and more faithful to long prompts than schnell.
No matches. Try “video”, “FLUX” or “webhooks”.
Loading models…

The open-weights FLUX workhorse, sharper and more faithful to long prompts than schnell.

Precise local edits from one reference image, such as changing colors, adding objects or rewriting text.
Black Forest Labs' video model, 5 to 20 second clips with sound at 720p or 1080p.
Turns one portrait and a voice track into a talking, gesturing avatar video.
Topaz Labs' image upscaler with face recovery, denoise and sharpening, up to 4x.
Clean, commercially safe background removal for products and people.
No models match these filters
Clear filtersSmall vision-language model that returns a text answer for a single image given by URL
moondream2Vision
≈ $0.01
Integrate
moondream2/object-detectionVision
$0.02 / image
Integrate
moondream2/point-object-detectionVision
$0.02 / image
Integrate
moondream2/visual-queryVision
$0.01 / 1000 characters
Integrate