Recursive ships Flint, a local coding model built with automated AI research
Richard Socher's first model at Recursive is open-weight and Apache 2.0 licensed, with performance and speed claims the company has not independently verified.
By Ryan Merket · Published
Primary source: Richard Socher on X
Why it matters
Flint gives Recursive's automated-research thesis a public, downloadable test case. The company's benchmarks and speed figures are self-reported, so developer results and repeat improvements across future models will determine whether the Eureka Machine is more than a compelling account of how Flint was built.
Richard Socher (@RichardSocher) introduced Flint on October 21st as the first model from Recursive, a coding model designed to run locally and built with the startup's automated research system, the Eureka Machine. The launch puts a developer-facing product behind Recursive's larger bet: using AI to propose, implement and test improvements to AI systems. Socher announced the release in a thread on X.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.