Google said on October 1, 2026, that its Planet-built Project Suncatcher prototype had launched into orbit aboard SpaceX’s Transporter-18 rideshare mission. The company reported contact with the satellite and said it was operating as expected. Its paper’s roughly 1,800 Starship launches, meanwhile, are part of a conditional launch-cost model—not a launch plan announced by SpaceX.
Project Suncatcher is an early research mission intended to collect data on how Google’s tensor processing units (TPUs), specialized chips for machine-learning tasks, respond to launch stress, radiation and thermal conditions in orbit. The satellite’s status marks the start of that research effort; the project’s broader aim is to explore orbital computing infrastructure.
What the 1,800-launch estimate assumes
Google’s paper models a possible path to launch prices below $200 per kilogram to low Earth orbit by about 2035. Its scenario depends on a learning rate of roughly 20%: the modeled price per kilogram falls by about that amount each time the cumulative mass launched doubles. That is a per-doubling assumption, not a yearly reduction.
The paper estimates that the scenario would require about 370,000 additional tonnes launched to orbit. It translates that mass into approximately 1,800 Starship flights, assuming 200 tonnes per flight, and a pace of about 180 launches a year.
| Model element | Figure | What it represents |
| Additional cumulative mass | About 370,000 tonnes | Mass launched in the modeled scenario |
| Assumed payload | 200 tonnes per flight | The paper’s assumed mass carried on each flight |
| Starship launches | About 1,800; about 1,850 by division | The paper’s rounded flight estimate and the quotient from its rounded mass and payload figures |
| Modeled cadence | About 180 launches per year | Annual pace associated with the scenario |
| Projected launch price | Below $200/kg by about 2035 | Conditional estimate for launches to low Earth orbit |
The rounded mass and payload figures work out to about 1,850 flights when divided; the paper gives the equivalent as approximately 1,800. A separate scenario puts the price around $300 per kilogram if 104,000 tonnes are launched by the mid-2030s. Google describes its cost analysis as a high-level evaluation rather than a comprehensive economic-feasibility study.
The proposed research path beyond the prototype
The paper uses an 81-satellite formation as an illustrative configuration for its analysis. Google’s stated plan for 2027 is to test two satellites in orbit and evaluate optical links between them—an important step toward studying how spacecraft in a proposed orbital system could communicate.
