Model card
MotionLab
MotionLab brings directed video generation to creative production through MotionLab Fast and MotionLab Max, with camera controls, flexible inputs, safety checks, and API delivery.
Current capabilities
- Model ID
- motionlab/fast
- Modes
- image_to_video, text_to_video, image_reference_video_to_video, image_audio_to_video, first_last_frame_to_video, seamless_loop
- Duration
- 3–20 seconds across enabled tiers
- Resolution limits
- 720p: 3–20s, 1080p: 3–20s
- Audio
- Generated soundtrack on by default; uploaded audio is supported; output audio can be disabled
- Output
- MP4 video over HTTPS; copy the output before its expiry
Intended use
Motion design, event and LED-screen visuals, social and marketing clips, concept animation from owned stills, and approved platform integrations. When enabled, Loop mode is designed to improve boundary continuity for repeating playback. Review the generated clip for your intended use.
Limitations
- Standard image-to-video requires a start image. Prompt-only generation is available only when live capabilities list text_to_video.
- Loop requires a start image and excludes an end image and Multi-Reference mode. A perceptually perfect loop is not guaranteed.
- Output quality depends on source quality and prompt specificity.
- Motion and subject consistency are probabilistic, not guaranteed.
- Availability is controlled by live feature flags and pricing configuration.
Safety and API lifecycle
Prompts and input images pass safety checks. The API uses scoped keys, idempotent job creation, signed webhooks, stable error codes, and automatic single refunds for qualifying technical failures or cancellation.