Open foundation AI models for world simulation
Video, audio & multimodal AI models built for enterprises, researchers, developers & studios.
.svg)
We build open world models that give you full control, from production-grade video to systems that understand and operate in the physical world.
Innovation
Research
Research
LTX-2.3 is built on a 22B-parameter asymmetric dual-stream diffusion transformer: a 14B video stream and 5B audio stream with bidirectional cross-attention. The full paper, weights, and code are publicly available.
Asymmetric dual-stream DiT architecture. Joint audio-video generation with bidirectional cross-attention and modality-aware classifier-free guidance. Open weights and code.
Rebuilt VAE for sharper detail. 4x larger text connector for tighter prompt adherence. Native portrait, cleaner audio, HDR output, and precise control over motion and camera.
For academic teams pushing the boundaries of video generation, world simulation, and multimodal AI. Grants, model access, and research partnerships with the LTX team.
WHAT'S NEW
Introducing the new LTX Trainer
Customize the model, not just the output. Securly train on your own characters, styles, workflows, and IP with the official open-source training framework for LTX.
New conditioning modes, flexible workflow composition, and an agentic setup experience — built to fit your pipeline, not the other way around. Learn More →
Our Stack
Built on LTX
Built on LTX
Complete flexibility in how you access and deploy.
Licensing
Full model weights for on-premise deployment and commercial use. No usage limits, no dependencies.
API
Integrate LTX-2.3 directly into your product or pipeline. Managed endpoints, no infrastructure overhead.
LTX Studio
A full production suite showing what the model enables for creative teams, studios, and enterprise.
Customer Voices
"LTX is a driving force in AI video, so when the opportunity came up to work together, it felt like a no-brainer. Audio driving video, where sound becomes the control layer is the first of its kind in AI video generation. So when we launched LTX's model within ElevenCreative and saw such a notable volume in video generations, it wasn't a surprise. Creators were ready, they just needed the tools to catch up."
“LTX's HDR IC-LoRA gives our compositors and colorists the latitude to truly blend these elements rather than just stack them. We're applying it across our generative animation and augmented crowds, and the line between what's generated and what's traditionally rendered disappears.”
"When editing AI-generated shots, the final edit always looks off, different contrast, values, color balance from shot to shot. We try to unify it in post, but 8-bit SDR leaves almost no latitude for color correction."
"The industry has long needed a bridge between generative AI and professional finishing standards. By moving past 8-bit SDR, we’ve eliminated the technical gap that kept AI assets from being used on high-fidelity displays and within complex spatial experiences. We’re no longer compromising on bit depth; we’re finally getting the dynamic range required for cinematic immersion in XR and virtual production. In the past, AI-generated content was a black box; you couldn't relight it or grade it without the image falling apart. Now, these assets behave like the real world, carrying the dynamic range needed to sit alongside traditionally captured elements. This gives our teams a professional-grade toolkit to integrate generative AI into their creative process."
