Google’s AI Satellite Transmits First Response From Orbit Under Project Suncatcher
Google booted TPUs aboard an orbiting satellite and ran Gemma 4, transmitting a space haiku as its first AI response. Next, Project Suncatcher will test thermal and radiation resilience ahead of planned twin satellites in 2027.
Google’s AI-equipped satellite has successfully generated and transmitted its first response from space, the company announced in a blog post on Friday, October 30, 2026.
The milestone involved booting up the satellite’s Tensor Processing Units (TPUs) in orbit and running the on-board Gemma 4 model. As an initial test, the model was prompted to write a haiku about being in space, producing: “Blue marble below / Data streams in silent grace / Starlight feeds my core. ✨”
Google says it also tested other space-themed limericks and telemetry prompts during the session. The satellite was launched into orbit earlier in October 2026.
The test is part of Google’s Project Suncatcher, an experimental initiative aimed at eventually placing AI data centers in space. Google has not yet detailed the full scope or timeline of the project beyond its next stated milestone.
“Next, we will continue with a series of tests to evaluate how TPUs perform under thermal stress and radiation exposure in orbit,” Google said. “Verifying our TPUs can reliably execute workloads in space paves the way for our next milestone of twin interconnected satellites in 2027.”
The successful transmission suggests that AI inference workloads may be viable in the harsh conditions of orbit, though Google’s planned thermal and radiation stress tests will be needed to confirm the hardware’s reliability over time. If those tests succeed, the project could eventually support a network of space-based AI processing nodes, with a two-satellite interconnected system targeted for 2027.