Skip to content

LTX-2.5 model license

Accepting the license

  • Log in to or create a Hugging Face account (free).
  • Open the gated Lightricks/LTX-2.5 model repository.
  • Read the LTX-2 Community License Agreement and the information on the access form.
  • Click Agree and Access. Access is normally granted automatically.
  • Create or use a Hugging Face token. A read token is sufficient; a fine-grained token must have read access to Lightricks/LTX-2.5 or to public gated repositories you can access.
  • Choose between the following options.

Access belongs to the individual Hugging Face account that accepted the terms. A token from another account will not inherit that access.

Option 1: Before starting the template

  • Click Edit Template and expand the Environment Variables section.
  • Add your token or RunPod secret to HF_TOKEN.
  • For a private template, the supplied environment profiles use:
HF_TOKEN={{ RUNPOD_SECRET_HF_TOKEN_WRITE }}
  • Create the HF_TOKEN_WRITE RunPod secret first, or change the reference to your existing secret name.
  • Save the template.
  • Click Deploy On-Demand.
  • The template will download the selected LTX-2.5 transformer, text encoder, distilled LoRA, VAEs, upscalers, and model patch.

Never save a real Hugging Face token in a public template.

Option 2: When the pod is running

Set your Hugging Face token

  • Open a terminal from the console or Code Server (see tutorial).
  • Fill in your Hugging Face token (HF_TOKEN).
export HF_TOKEN=Your-hf-token-here

or

hf auth login
  • Press enter.
  • Verify the connected account and gated repository access.
hf auth whoami
hf download Lightricks/LTX-2.5 \
  diffusion_models/ltx-2.5-22b-dev-transformer-comfy-int8-convrot.safetensors \
  --dry-run
  • Manual provision from the documentation available in the pod
  • Press enter.
  • Wait until the download finishes.
  • Restart the pod if the startup provisioning previously failed; existing files will be skipped and the remaining chain will be downloaded.