LTX-2.5 inference¶

FastVideo supports BF16 inference for the LTX-2.5 dev and distilled transformers. Both text-to-video and image-to-video generate synchronized video and audio. The native path supports sequence parallelism, component offload, and torch.compile.

Convert the official checkpoint¶

The gated Lightricks/LTX-2.5 repository publishes separate transformer, packed Gemma 4, convolutional video VAE, audio VAE/vocoder, and spatial upscaler files. Accept its license and download those files, then convert them into one FastVideo model directory:

python scripts/checkpoint_conversion/convert_ltx2_weights.py \
  --variant distilled \
  --transformer-source /weights/ltx-2.5-22b-distilled-transformer-bf16.safetensors \
  --text-encoder-source /weights/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors \
  --vae-source /weights/ltx-2.5-video-vae-conv-bf16.safetensors \
  --audio-vae-source /weights/ltx-2.5-audio-vae-bf16.safetensors \
  --spatial-upscaler-source /weights/ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors \
  --output /models/LTX-2.5-Distilled-Diffusers

Use --variant dev and the dev transformer to convert the development model. The converter emits a standard component directory and preserves LTX-2.5's architecture metadata, packed tokenizer, joint audio components, and refine configuration. The convolutional video VAE is the supported launch path.

Generate video and audio¶

Run distilled text-to-audio-video across four GPUs with sequence parallelism:

python examples/inference/basic/basic_ltx2_5_t2av.py \
  --model-path /models/LTX-2.5-Distilled-Diffusers \
  --prompt "A jazz trio performs in a candlelit club, synchronized live sound" \
  --num-gpus 4 \
  --torch-compile

Condition on a first frame with the I2AV example:

python examples/inference/basic/basic_ltx2_5_i2av.py \
  --model-path /models/LTX-2.5-Distilled-Diffusers \
  --image /images/first-frame.png \
  --prompt "The train pulls away from the platform as its horn sounds" \
  --num-gpus 4 \
  --torch-compile

Pass --variant dev with a converted dev directory to use the 30-step dev guidance preset. Distilled inference uses the official eight-step ancestral schedule and a three-step spatial refinement pass by default.

Current scope¶

The initial inference path includes the native transformer, packed Gemma 4 text stack, convolutional video VAE, audio VAE/vocoder, dev guidance, and the distilled ancestral sampler. DiffVAE/NATTEN, generated keyframes, temporal upsampling, automatic duration selection, HDR, training, fine-tuning, and quantized deployment are separate follow-up work.