Understand this story
On September 4, 2026, Markov Robotics founder Diego Caples said that Markov Robotics uses Modal for almost all of its computing. In an official engineering blog post the same day, Modal introduced that the company runs all of its preprocessing, training, and deployment on Modal, and that it trained a model capable of performing a variety of complex manipulation tasks using less than ten hours of robot data. Its preprocessing pipeline, with the help of .spawn_map, scales to over a thousand containers and up to 500 GPUs, shortening a task that originally took a full day to 20–30 minutes. The team runs 2 to 10 training jobs simultaneously and uses Modal Tunnels to connect the robots directly to GPUs in us-west, with round-trip network overhead of about 20 milliseconds.
AI-generated from public reports
Latest development · Report
Markov Robotics Uses Modal to Complete End-to-End Training on a Single PlatformDevelopments
1 confirmed developments
Markov Robotics founder says the entire pipeline of computation runs on Modal
Source:English text is being prepared.
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Sep 4
Modal Official Engineering BlogFirst-party
Markov Robotics Uses Modal to Complete End-to-End Training on a Single Platform
Markov Robotics runs preprocessing, training, and deployment entirely on Modal, training a model capable of executing a variety of complex manipulation tasks from less than ten hours of robot data. Its preprocessing pipeline uses .spawn_map to scale to over a thousand containers and up to 500 GPUs, shortening a task that originally took a full day to 20–30 minutes. The team runs 2 to 10 training jobs simultaneously and connects the robots directly to GPUs in us-west via Modal Tunnels, with round-trip network overhead of about 20 milliseconds.