From the model card
Fields below are copied from the tags and counters on the HuggingFace repository DeepBeepMeep/LTX-2 at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- DeepBeepMeep
- Weight formats
- ONNX, safetensors, GGUF
- Lineage
-
- base model Lightricks/LTX-2
- quantized from Lightricks/LTX-2
- Downloads (HF counter at last fetch)
- 416,627
- Likes (HF counter at last fetch)
- 25
- Model card
- https://huggingface.co/DeepBeepMeep/LTX-2
Use cases
- Short video clip generation from text prompts
- Video generation experimentation on consumer GPUs
- Creative video content prototyping
- Research into DiT-based video generation architectures
Pros
- LTX Video lineage is optimized for faster generation than Wan or CogVideoX
- Community fine-tune may improve quality on specific styles or content types
- Open weights enable local video generation without API costs
Cons
- Limited documentation on what LTX-2 specifically fine-tunes or improves
- Video generation quality lags behind commercial services
- High compute requirement even for short clips
- Community fine-tune provenance and training details not disclosed
Tags
diffusion-single-fileonnxsafetensorsggufbase_model:Lightricks/LTX-2base_model:quantized:Lightricks/LTX-2region:us