Tasks and inputs
LongCat Video Avatar 1.5 is listed for Image-to-video, Text-to-video, Audio-Driven Video. It accepts text, image, audio and produces video. Use the publisher’s input format and preprocessing for your chosen deployment.
SyncDev model directory · Profile updated Oct 8, 2026
LongCat Video Avatar 1.5 generates audio-driven character video from speech, scene instructions and optional image references.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
LongCat Video Avatar 1.5 generates audio-driven character video from speech, scene instructions and optional image references.
Avatar 1.5 builds on the LongCat Video framework for audio-text-to-video, audio-text-image-to-video and video continuation. The supplied audio drives a character performance, while a prompt and optional reference image describe appearance, action and setting.
The release includes single-stream and multi-stream audio examples. Its merge mode expects clips of equal length and combines them; its concatenation mode sequences clips and can add silence between turns. Choose the mode to match the scene instead of assuming every multi-speaker request has the same timing behaviour.
Use a short line, a longer passage and a scene with a second speaker. Check lip synchronization, identity preservation, body motion and transitions at continuation boundaries. For singing, review the timing of the visible performance against the supplied audio. References and prompts should describe the same subject consistently.
The repository documents distributed inference, distilled settings and an optional INT8 path for the Avatar 1.5 model. Hardware demand depends on duration, resolution and configuration. The weights use MIT licensing, while audio and image references still need to be suitable for the intended use. This checkpoint does not establish a per-second managed API price.
Compare the model on the work you need to complete.
LongCat Video Avatar 1.5 is listed for Image-to-video, Text-to-video, Audio-Driven Video. It accepts text, image, audio and produces video. Use the publisher’s input format and preprocessing for your chosen deployment.
Choose duration, resolution and frame rate together. Compare the cost of an accepted clip, including repeat generations.
The weights are downloadable under MIT. Check runtime support and size the complete model and working context for your expected concurrency.