FLUX 3 Max Image-to-Video
#1 image-to-video quality. Over 25× faster generation.
FLUX 3 Max delivers frontier image-to-video quality with substantially lower generation latency. Post-trained by fal Research in collaboration with Black Forest Labs, it ranks joint #1 in the preliminary Artificial Analysis image-to-video evaluation, statistically alongside the highest-ranked models.
In benchmarks, FLUX 3 Max generates a 10-second video in approximately 11 seconds, compared with 277.8 seconds for Seedance 2.5. That is more than 25× faster generation, pairing competitive visual quality with shorter waits between creative iterations.
The Image-to-Video endpoint turns a starting image into a moving scene. The image establishes the opening composition, while a text prompt directs subject movement, camera motion and changes within the scene. An optional ending image guides the final frame.
Key features
- Frontier visual quality: Joint #1 in the reported preliminary image-to-video evaluation.
- Optimized inference: fal Research’s post-training and inference optimization improve on the original FLUX 3 in both generation quality and speed.
- Image-conditioned generation: Begin with a photograph, illustration or generated image.
- Text-directed motion: Describe how the subject, camera and environment move after the opening frame.
- Optional ending frame: Supply a second image to guide where the sequence finishes.
- 5 to 20-second clips: Select the duration in one-second increments.
- 720p and 1080p output: Generate videos at 720p by default, with optional 1080p upscaling for higher-resolution output.
- Optional generated audio: Enable or disable audio generation.
Generation controls
| Control | Supported values |
|---|
| Required inputs | Starting image and text prompt |
| Optional image input | Ending frame |
| Image formats | PNG, JPEG, WebP |
| Duration | 5–20 seconds; 5 seconds by default |
| Resolution | 720p or 1080p; 720p by default |
| Aspect ratio | Auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16 |
| Automatic framing | Follows the starting image |
| Audio generation | Enabled by default; optional |
| Reasoning level | Low or high; high by default |
| Seed | Optional; randomly selected when omitted |
Directing motion from an image
Motion prompts describe what changes after the opening frame. Subject actions, camera movement and environmental details provide direction without requiring the prompt to repeat every visible element in the source image.
The camera slowly moves closer as the woman turns toward the window. A light breeze moves the curtain and a few loose strands of hair. She pauses, smiles slightly and looks outside. Quiet room ambience and distant street sounds.
This endpoint supports animated portraits, product scenes, illustrated sequences and motion studies that begin with an established visual composition.
About the benchmarks
The joint #1 result describes a preliminary image-to-video evaluation. Rankings can change as evaluations receive more votes. The latency comparison describes the tested benchmark configuration, rather than a guaranteed completion time for every request. Actual generation time varies with request settings and operating conditions.
More ways to create with FLUX 3 Max
Text-to-Video creates a scene without a starting image. Keyframes-to-Video provides image guidance at multiple points throughout a sequence.
Image-to-Video API documentation
blackforestlabs/flux-3-max/image-to-video
