LongCat Video Avatar: Audio-Driven Talking Head

A LongCat video avatar is not a cartoon filter or a sticker pack. It is an audio-driven talking head: you feed it a still portrait and a short voice clip, and LongCat generates a video where that face speaks the audio with matching lip movement. For faceless creators, virtual presenters, and anyone who wants a recurring on-screen persona without a camera, this is the difference between a static brand and a host who can deliver every script.

What "avatar" means here

In the LongCat ecosystem, the avatar workflow is a distinct inference path, not a prompt trick applied to the base video model:

• Audio image-to-video — a portrait plus a WAV file in, a speaking head video out
• Lip sync driven by the audio encoder, not a separate retiming pass after generation
• A dedicated checkpoint — LongCat-Video-Avatar-1.5 — whose audio encoder and inference path differ from the standard LongCat Video model
• Single-speaker friendly — clean, isolated speech produces the most stable mouth and head motion

The practical rule: treat the avatar model and the base video model as two different tools that happen to share a lineage. Swapping their checkpoints "because they are both LongCat" is the fastest way to a broken run.

Inputs that prevent wasted runs

Bad inputs cause more avatar failures than weak hardware. Before you generate anything, satisfy this checklist:

1. Use a sharp portrait with the full head visible and no object covering the face
2. Start with one speaker and a short, clean WAV file before attempting multi-person audio
3. Keep the audio sample rate and channel layout consistent across every test
4. Write a prompt that describes identity, framing, clothing, background, and restrained motion
5. Record the model filename, wrapper commit, workflow JSON, seed, dimensions, duration, and audio filename with every saved result
6. Confirm the audio has no loud background music or overlapping voices on the first pass
7. Avoid extreme close-ups that crop the jaw or forehead, which the lip-sync motion relies on
8. Reuse the same portrait across comparisons so you can isolate audio and prompt changes

Install the official path

When reproducibility matters more than convenience, use the official repository rather than a random re-uploaded workflow:

git clone --single-branch --branch main https://github.com/meituan-longcat/LongCat-Video
cd LongCat-Video

conda create -n longcat-video python=3.10
conda activate longcat-video

# Install the PyTorch/CUDA combination for your machine, then:
pip install ninja psutil packaging
pip install flash_attn==2.7.4.post1
pip install -r requirements.txt
conda install -c conda-forge librosa ffmpeg
pip install -r requirements_avatar.txt

Download the LongCat-Video-Avatar-1.5 weights linked by the repository. Do not silently substitute the original Avatar checkpoint: the audio encoder and inference path differ, and mixing files produces a checkpoint that will not load.

The ComfyUI route

For a visual workflow, the community wrapper keeps LongCat nodes on its main branch:

cd ComfyUI/custom_nodes
git clone https://github.com/kijai/ComfyUI-WanVideoWrapper.git
cd ComfyUI-WanVideoWrapper
pip install -r requirements.txt
git rev-parse HEAD

Record the commit printed by that last command with every saved workflow, then start from the repository's LongCatAvatar_audio_image_to_video_example_01.json example instead of an untraceable JSON from a video description. The wrapper is a personal, perpetually work-in-progress sandbox — update it in a separate copy, run one known input, and keep the last working commit so a node rename cannot break all jobs at once.

VRAM and duration: measure, do not guess

The official Avatar 1.5 pages do not publish a universal "works on 8GB / 12GB" minimum. Community FP8, GGUF, CPU offload, and block-swap workflows can reduce peak VRAM, but the real number depends on checkpoint format, resolution, frame count, attention backend, system RAM, and wrapper version. Use a bounded test ladder:

1. Run the repository's known example at its documented settings first
2. Run 3–5 seconds with your own portrait and audio
3. Record peak VRAM, system RAM, wall time, and output dimensions
4. Change only one variable at a time between runs
5. Increase duration only after lip sync and identity remain stable

"Supports long video" does not mean an infinite clip has zero drift. Continuation is a repeated generation operation — inspect every join for face changes, color shifts, hand artifacts, background jumps, and audio offset.

Failure triage

Production checklist before you publish

• Watch the complete result with sound and verify lip timing, identity, hands, teeth, and eye motion
• Check scene continuity at every continuation join
• Confirm you have the rights to the portrait and the audio you used
• Review the model license separately from "open source" — open weights do not equal commercial rights
• Apply any disclosure or consent requirements your platform imposes on synthetic media
• Keep the wrapper commit and model filename beside the final file for future reproducibility

Related use cases

• Consistent character AI video — keep the same face across long scenes
• Image to long video AI — animate a still image into minutes of motion
• LongCat AI video — the text-to-video and continuation engine underneath

For a faster first test, open the hosted image-to-video workflow.