gen‑ai.news
← Back
Multimodal

Reka AI's omni-model Rho-1 handles text, images, video, and robot control in a single model

Reka AI's omni-model Rho-1 handles text, images, video, and robot control in a single model

Reka AI has introduced Rho-1, a 19-billion-parameter omni-model designed to handle text, images, video, and robot control actions within a single unified neural network. The core architectural idea is straightforward: instead of building separate specialized models for each modality and coordinating between them, Rho-1 treats every input and output - whether a word, an image patch, a video frame, or a robot action - as tokens processed in one shared context window.

This unified approach stands in contrast to the more common practice of assembling multimodal systems from distinct components, such as a vision encoder paired with a language model, or a separate policy network for robotics. By collapsing everything into a single token stream, Rho-1 can in principle draw on relationships across modalities that would otherwise be lost at the handoff points between specialized modules.

On the compute side, Reka reports that Rho-1 was trained on 320 NVIDIA H100 GPUs over approximately three months. That is a meaningful data point in context - many frontier models today are trained on thousands of GPUs for longer periods. Whether the efficiency gains come from architectural choices, data curation, or training methodology, the relatively modest resource footprint suggests that omni-model approaches may not necessarily require the largest compute budgets to cover a broad range of tasks.

The inclusion of robot control as a native output modality is notable. Most robotics work still relies on separate perception and planning stacks, often with language models bolted on for instruction following. A model that generates control actions in the same framework it uses to reason about images and text could simplify deployment pipelines and potentially improve coherence between perception and action. Rho-1 represents an early indication of where this line of research is heading, though real-world robotics performance will depend heavily on how well the model generalizes beyond its training distribution.

Also covered by

Enjoy this story? Get the next one in your inbox.

Twice a week: the most important stories in generative image and video AI, distilled into a 2-minute read.

Free. Unsubscribe any time. No spam, ever.

Your next read

Suno now lets you make AI videos for your AI music
Multimodal

Suno now lets you make AI videos for your AI music

Suno has added a new "Visuals" feature that lets users generate AI video clips and images to accompany their AI-made music. Videos can automatically lipsync to song lyrics, while static images can serve as album artwork or thumbnails. The feature is available now across Suno's web, iOS, and Android apps.

Microsoft launches seven MAI models at Ignite 2026
Multimodal

Microsoft launches seven MAI models at Ignite 2026

Microsoft unveiled seven new models under its MAI lineup at Ignite 2026, spanning code generation, cybersecurity, image synthesis, real-time processing, and decision-making. The release signals a broader push by Microsoft to build out its own model portfolio alongside its existing partnership with OpenAI. Notable entries include MAI Code 2, aimed at GitHub Copilot, and MAI Image 2.7 for generative image tasks.

Building Visual Intelligence — Andreas Blattmann & Dustin Podell, Black Forest Labs
Multimodal

Building Visual Intelligence — Andreas Blattmann & Dustin Podell, Black Forest Labs

Andreas Blattmann and Dustin Podell of Black Forest Labs lay out their thinking on how generative models trained on visual data - pixels, motion, and sound - could serve as a meaningful path toward general intelligence. The conversation digs into the technical and philosophical foundations behind the lab's approach to building visual AI systems. It offers a grounded look at why the founders believe perception-driven learning is worth pursuing at the frontier.