Find what you need to get started with LTX

LTX is Lightricks' AI video generation platform — built around LTX-2.3, an open-source foundation model for cinematic video generation. It supports text-to-video, image-to-video, audio-to-video, and more, at up to 4K resolution. You can access it through the LTX API, LTX Studio, or LTX Desktop.

LTX builds open AI models for video and visual generation — the foundation for the next generation of creative production and world simulation. Anyone can use it: developers building AI-powered products, studios running production pipelines, or individuals exploring what the model can do. We have a few access points depending on who you are and what you're building — explore your options on ltx.io.

Yes. LTX-2.3 model weights, code, and tooling are publicly available on Hugging Face and GitHub. The open weights are free to use for most purposes — if your company generates over $10M in annual revenue, a commercial license is required. See the LTX Model License for details.

We have a few interfaces to access LTX, depending on who you are and what you're building. At the core are our open model weights — available for self-hosted, custom, or research deployments with full control over how you run and extend the model. From there, the LTX API lets developers integrate video generation directly into products and pipelines. LTX Studio is a browser-based platform for filmmakers and creative teams who want a ready-to-use interface. And LTX Desktop runs the full model locally on your own machine, no cloud required.

It depends on how technical you are and how much control you need. If you want to generate and explore without any setup, start with LTX Studio. If you're building a product or need programmatic access, the LTX API is your starting point. If you want full local control with no cloud dependency, LTX Desktop runs entirely on your machine. For custom training, self-hosting, or deep infrastructure integration, start with the open model weights.

Most AI video tools are closed, cloud-only, and API-locked. LTX is different in three ways: the underlying model is open source; you can run it locally or on your own infrastructure; and it's designed for production workloads, not just demos. That makes it a practical choice for teams with strict data, compliance, or customization requirements.

The fastest way to try the model is the LTX Playground — no setup required. From there, explore the product pages above to find the right entry point for your use case.