Key Info

Recursive SI (@Recursive_SI) announced Flint, its first model, described as a fast local coding model built with the company's automated research system, the Eureka Machine. Open weights are available on Hugging Face.

Highlights

  • Positioned as best-in-class fast local coding, claimed up to 2x faster than comparable models
  • Reported throughput of 125+ tokens/sec on a MacBook Pro and 150+ on RTX Spark
  • Built from scratch in under 4 months
  • Marketed as private, unmetered intelligence on the user's own machine
  • Can be run locally via popular frameworks, or tried free on OpenCode and Pidotdev