Tasks and inputs
LTX 2.5 is listed for Image-to-video, Text-to-video, Video-to-video. It accepts text, image, video, audio and produces video, audio. Use the publisher’s input format and preprocessing for your chosen deployment.
LTX 2.5 is Lightricks’ downloadable model for synchronized video and audio generation in a local production pipeline.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
LTX 2.5 is Lightricks’ downloadable model for synchronized video and audio generation in a local production pipeline.
The publisher provides a diffusion-transformer model with synchronized audio and video, a Python pipeline and ComfyUI integration. The documented workflow supports conditioning from text and reference media, along with local customization using supported adapters. The checkpoint is separate from the managed Fast and Pro API routes and their operation tables.
Start with a short scene whose action, camera direction and sound are easy to assess. Then test an image-conditioned shot and a scene requiring a transition between planned moments. Review continuity, subject identity and audio alignment throughout the clip. Keep the input files, adapter choices and generation settings with the result so an editor can reproduce a revision.
The public checkpoint metadata identifies the LTX 2.x Community License Agreement. Access to the model files requires the publisher’s acceptance flow; the directory does not accept those terms on the reader’s behalf. Follow the official local setup and review the checkpoint and code terms before deployment. Do not transfer an Apache licence from an unrelated video model.
Measure the complete local pipeline, including model loading, preprocessing, generation and output encoding. Hardware, precision and output settings affect the result, so this profile does not infer a generation speed or hosted token price from download availability. Compare a managed route separately if the team prefers a maintained API to operating the local pipeline.
Publisher documentation · Publisher documentation · Publisher documentation
Compare the model on the work you need to complete.
LTX 2.5 is listed for Image-to-video, Text-to-video, Video-to-video. It accepts text, image, video, audio and produces video, audio. Use the publisher’s input format and preprocessing for your chosen deployment.
Choose duration, resolution and frame rate together. Compare the cost of an accepted clip, including repeat generations.
The weights are downloadable under LTX 2.x Community License Agreement. Check runtime support and size the complete model and working context for your expected concurrency.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
MiniMax
MiniMax H3 is a multimodal generation system that produces video with synchronized stereo audio from supported text, image, video and audio references.
ByteDance Seed
Seedance 2.5 from ByteDance Seed: ByteDance multimodal audio-video model with up to 30-second single-pass generation, multi-round extension, reference inputs and timestamp-level editing.
Gemini Omni Flash Preview from Google: Video generation and editing model for fast, conversational text- and image-to-video workflows.
xAI
Grok Imagine Video 1.5 creates video from text, image references and supported voice inputs, with frame controls for planned shots.
Veo 3.1 Lite Preview from Google: Video model for prompt-guided generation, editing, and motion workflows.
ByteDance Seed
Seedance 2.0 from ByteDance Seed: ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.