Longcat Full Quality
Meituan's 13.6B open-source video model — generating coherent, high-quality private clips up to 30 seconds on Venice.
Overview
What is Longcat Full Quality
Longcat Full Quality is an open-source, 13.6-billion-parameter video generation model developed by Meituan. Built on a Diffusion Transformer (DiT) architecture, it excels at producing highly coherent, high-quality videos. On Venice, it supports text-to-video generation at 720p resolution with flexible aspect ratios and clip lengths up to 30 seconds.
Running it privately on Venice
On Venice, you can generate videos with Longcat completely privately under our zero-retention policy, ensuring your creative prompts remain entirely your own. Venice offers an uncensored environment for this open-weights model, allowing unrestricted creative freedom. Instead of expensive subscriptions, you pay only per clip, making high-fidelity video generation highly accessible.
Assessment
Strengths and limitations
- Strong temporal coherence and visual consistency across longer video durations.
- Fully open-weights model under the permissive MIT license, ensuring user sovereignty.
- Unified Diffusion Transformer (DiT) architecture optimized with Block Sparse Attention for efficient inference.
- Trained with multi-reward RLHF (GRPO) for superior prompt alignment and natural motion.
- No audio generation support natively on Venice.
- Maximum resolution is currently capped at 720p on the platform.
- Can be computationally intensive to run locally compared to smaller video models.
Capabilities
What it supports
- Text to video
- Image to video
- Reference to video
- Native audio generation
Variants
Longcat Full Quality model variants
Longcat Full Quality runs on Venice as 2 variants of the same underlying model. Pick by what you're starting from: a written prompt, a still image, reference images, or an existing clip. Each variant is its own model id on the API; the generation quality is the same across the family.
| Variant | What it is | Clip lengths | Resolutions | Aspect ratios | Audio | Model ID |
|---|---|---|---|---|---|---|
| Text to Videoflagship | Generate a clip from a written prompt | 5s – 30s | 720p | 16:9, 9:16, 1:1 | longcat-text-to-video | |
| Image to Video | Animate a still image into motion | 5s – 30s | 720p | — | longcat-image-to-video |
Capability data comes straight from the Venice model API and refreshes with every catalog ingest. The specs and pricing on this page are captured from the flagship variant; pass the model id of the variant you want to the API.
Longcat Full Quality Text to Video
Generate a clip from a written prompt. Supports clips of 5s – 30s, 720p output, 16:9, 9:16, 1:1 aspect ratios.
longcat-text-to-videoLongcat Full Quality Image to Video
Animate a still image into motion. Supports clips of 5s – 30s, 720p output.
longcat-image-to-videoSpecifications
Datasheet
- Maker
- Meituan
- Released
- October 2025
- Architecture
- Diffusion Transformer (DiT)
- Parameters
- 13.6 Billion
- Open weights
- Yes (MIT License)
- Resolutions
- 720p
- Clip lengths
- 5s, 10s, 15s, 20s, 30s
- Mode
- text-to-video
- Aspect ratios
- 16:9, 9:16, 1:1
- Audio
- No
- Privacy on Venice
- Private — zero retention
- Available on Venice since
- Dec 2025
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id.
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "longcat-text-to-video",
"prompt": "Aerial drone shot over a misty mountain valley at golden hour"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Pricing
What it costs on Venice
Pay per clip on Venice — price scales with resolution and duration (5s–30s), from $0.25.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
Alternatives
How it compares
| Model | Best for | Max resolution | Open weights | Uncensored | Price (Venice) |
|---|---|---|---|---|---|
| Longcat Full Quality | Excellent temporal coherence and open-source flexibility. | 720p | Yes | Yes | from $0.25 |
| Wan 2.7 | A powerful alternative open-weights video model. | 720p | Yes | Yes | from $0.55 |
| Grok Imagine Private | Proprietary model with private inference options. | — | No | Yes | from $0.32 |
Excellent temporal coherence and open-source flexibility.
Use cases
What it is good for
- 01Creating cinematic B-roll and atmospheric video clips from text descriptions.
- 02Generating social media video assets in 9:16, 16:9, or 1:1 aspect ratios.
- 03Prototyping visual concepts and storyboards privately without subscription lock-in.
Prompting
Getting better results
Be descriptive with motion and lighting cues to leverage the DiT architecture's spatial understanding.
Choose the 9:16 aspect ratio for mobile-first platforms, or 16:9 for cinematic widescreen outputs.
Start with shorter 5-second clips to test your prompt composition before committing to longer 30-second generations.
Version history
Initial release of the 13.6B parameter model with unified video generation capabilities.
FAQ
Frequently asked questions
Longcat Full Quality is an open-source, 13.6-billion-parameter video generation model developed by Meituan. It uses a Diffusion Transformer (DiT) architecture to generate high-quality, coherent video clips from text prompts.
On Venice, you pay per clip starting from $0.25 for a 720p, 5-second video. The price scales with the selected clip duration (up to 30 seconds).
Yes, Longcat is fully open-source under the MIT license. On Venice, you can try it using free daily account credits, or pay per clip for heavier usage.
No, Longcat currently generates silent video clips. You will need to add audio or sound effects in post-production.
Longcat supports three main aspect ratios on Venice: 16:9 (widescreen), 9:16 (vertical/mobile), and 1:1 (square).
Longcat excels at temporal coherence and is natively pretrained for video continuation, making it highly stable. Wan 2.7 is another excellent open-weights alternative with distinct aesthetic qualities.
Yes. Venice operates under a strict zero-retention policy. Your prompts and generated videos are never stored, profiled, or used to train models.
Yes. Alongside the text-to-video model (longcat-text-to-video), Venice hosts longcat-image-to-video, which animates a still image into a moving clip. Both variants render at 720p, support clip durations of 5, 10, 15, 20, or 30 seconds, and run privately with zero prompt retention.
Run Longcat Full Quality privately
No prompt logging. No data used for training.