LTX-2.5 路 distilled 路 bf16 (ungated mirror for NVIDIA GPUs)

An unmodified, ungated copy of the files needed for the fast DistilledPipeline from Lightricks/LTX-2.5 (revision 49b093a39c19169a8a81721dcc541e07e945232f), so they can be downloaded without a Hugging Face account.

Credits

  • Model: LTX-2.5 by Lightricks.
  • This mirror only re-hosts their files. It is not affiliated with or endorsed by Lightricks.

What is in this copy

  • Every file is byte-identical to the source revision above (SHA-256 checked before uploading), kept in the same folder layout. No file was modified.
  • Only the 5 files of the README's DistilledPipeline command (about 66 GiB):
    • diffusion_models/ltx-2.5-22b-distilled-transformer-bf16.safetensors
    • text_encoders/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors
    • vae/ltx-2.5-video-vae-bf16.safetensors
    • vae/ltx-2.5-audio-vae-bf16.safetensors
    • latent_upscale_models/ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors
  • Everything else in the source repository (dev transformer, int8/nvfp4 variants, LoRAs, temporal upscaler) was omitted.
  • Runs with the official pipeline: python -m ltx_pipelines.distilled from github.com/Lightricks/LTX-2.

License

These weights are distributed under the LTX-2.x Community License Agreement. A verbatim copy is in LICENSE; the upstream text is at https://github.com/Lightricks/LTX-2/blob/main/LICENSE-2_x.

By downloading or using these weights you agree to that license, including the use restrictions in its Attachment A. If you redistribute them, you must pass the same license and restrictions along.

Used by

VideoMakerIA, a local video studio, which shows the full license text and asks for acceptance before downloading these files.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 馃檵 Ask for provider support

Model tree for MarceloFSO/ltx-2.5-distilled-nvidia

Finetuned
(34)
this model