Loft Orbital's Yam-9 Runs Google DeepMind's Gemma 3 Directly in Space
NASA's JPL ran Google's Gemma 3 vision-language model aboard a Loft Orbital satellite, marking the first time a VLM has ever analyzed imagery autonomously in space.
- NASA JPL's NAVI-Orbital software ran Google DeepMind's Gemma 3 VLM aboard Loft Orbital's Yam-9 satellite — the first vision-language model ever deployed in orbit.
- The satellite autonomously answered natural language queries and classified Earth imagery without downlinking raw data to the ground first.
- The hardware stack: an Nvidia Jetson Orin AGX GPU running an off-the-shelf Gemma 3 model, adapted by JPL engineers to fit satellite memory constraints.
- Use cases include border monitoring, railway infrastructure detection, agricultural surveillance, and eventually AI assistants for astronauts on the Moon or Mars.
- Loft Orbital has a defense deal with Helsing for real-time military situational awareness; full Earth coverage would require 50–100 satellites (Loft currently operates 12).
- Kepler Communications is running 40 Nvidia GPUs across 10 satellites in orbit, with undisclosed VLM use cases already underway.
For decades, Earth observation satellites have operated on a simple and expensive loop: take photos, beam them down, and let humans sort through the mess. That loop just broke. In April, Loft Orbital's Yam-9 satellite became the first spacecraft to autonomously identify targets in orbit using a vision-language model (VLM) , a type of AI that combines image understanding with the reasoning of a large language model. No analyst in the loop. No raw data downlinked first. The satellite just... figured it out.
What actually happened up there
The milestone occurred in April and marks the first reported use of a vision-language model in orbit. The Yam-9 satellite ran a software package from NASA's Jet Propulsion Laboratory called NAVI-Orbital, which hosted Google DeepMind's Gemma 3 VLM. Aboard the spacecraft, Gemma 3 answered natural language queries, classified Earth imagery, and produced plain-English summaries of what it found , without transmitting a single raw pixel to the ground first.
A VLM, for context, is a model that can take both an image and a text prompt as input and reason about them together. Think of it as giving a satellite eyes and the ability to understand instructions like "find me railway infrastructure" rather than just capturing everything and hoping someone on the ground notices what matters.
Researchers asked the model to classify sensor data where natural environment meets human development, or to detect infrastructure around railway hubs , and it handled it. The team used LangGraph for agent orchestration and Gemma 3 to generate short English summaries right from space.
The stack that made it fly
This wasn't a custom space-grade AI built from scratch. The engineering challenge was about compression and adaptation, not invention.
- Model: Gemma 3 from Google DeepMind, an open-weights VLM available in sizes from 1B to 27B parameters, purpose-built for edge deployment