Decart
Today, we are releasing MirageLSD, the first Live-Stream Diffusion (LSD) AI video model that inserts your imagination into any live video feed — it is the difference between watching magic on a screen and making magic yourself. We are also showing a demo of Mirage, our platform (website today, iOS/Android coming next week), powered by the MirageLSD model, to change live streaming, video calls, the camera, VOD, gaming and more. Unlike other AI video models with 10+ second delays and 5-10s clips, Mirage transforms infinite video streams in realtime (<40ms response time). Mirage is our second model after Oasis (the viral "AI Minecraft"), took roughly half a year to build, and required solving cutting-edge math challenges and writing efficient GPU assembly code. For the technical deep dive, check out the details below. Mirage is day one. Next come a variety of additional video models, audio, emotions, music—we're building the platform that turns all our senses into portals. Today, you can
MirageLSD: The First Live-Stream Diffusion AI Video Model | Decart AI {% if %} {% endif %} Oasis 3 is live - the first API-accessible world model for Physical AI → contact us July 17, 2025 MirageLSD: The First Live-Stream Diffusion AI Video Model MirageLSD is the first system to achieve infinite, real-time video generation with zero latency. Try Mirage Today, we are releasing MirageLSD, the first Live-Stream Diffusion (LSD) AI video model that inserts your imagination into any live video feed — it is the difference between watching magic on a screen and making magic yourself . We are also show
saved by
related reading
- Video Generation Models Explosion 2024yenchenlin.me
- How do AI models generate videos? | MIT Technology Reviewtechnologyreview.com
- What are Diffusion Models? | Lil'Loglilianweng.github.io
- Sand.aisand.ai
- Flexible Diffusion Modeling of Long Videosarxiv.org
- What are Diffusion Models?lilianweng.github.io
- ⭐️ Diffusion Modelsandrewkchan.dev
- A Dive into Text-to-Video Modelshuggingface.co
- RTFM: A Real-Time Frame Model | World Labsworldlabs.ai
- Kuleshov Group | How to Build a Diffusion Language Modelkuleshov-group.github.io
- Video models are zero-shot learners and reasonersarxiv.org
- Replicate - Run AI with an APIreplicate.com