Recursive Releases Flint, a Free Laptop Coding Model Built Largely by Its Own AI System
Flint runs locally at about 131 tokens per second on an M4 Pro MacBook—roughly twice the speed of comparable models. The 40B-parameter MoE activates 2.5B parameters, is free under Apache 2.0, and works offline for private coding.
AI startup Recursive released Flint on October 21, 2026 — a free, laptop-sized coding model the company says was built largely by its own automated AI research system, called the Eureka Machine.
Flint is a mixture-of-experts model with 40 billion parameters that activates only 2.5 billion at a time, making it light enough to run on an ordinary laptop. Because it runs locally rather than on cloud servers, developers can use it offline, keep their code private, and avoid usage fees. The model is available for free download on Hugging Face under the Apache 2.0 open-source license.
Recursive says the Eureka Machine contributed to nearly every stage of Flint’s development — from pre-training to post-training and inference — allowing a small team to build the model from scratch in under four months. The system’s contributions include a sparser model architecture, low-bit training that reduced Flint’s size by 71% with minimal quality loss, and custom computational kernels. Recursive did not specify how much time the Eureka Machine saved its researchers.
In the company’s tests, Flint generated around 131 tokens per second on a MacBook Pro with Apple’s M4 Pro chip — roughly twice the speed of comparable models including Google’s Gemma 4 and Nvidia’s Nemotron 3.5 Lightning on the same hardware. On SWE-Bench Verified, a benchmark for fixing real-world software bugs, Flint outperformed all other models under 50 billion parameters. Across Recursive’s own six-test average, only Alibaba’s Qwen3.8 27B scored higher.
Recursive, founded in 2025, emerged from stealth in May 2026 with $650 million in funding at a $4.65 billion valuation, backed by GV, Nvidia, and AMD Ventures. The company is led by CEO Richard Socher, formerly chief scientist at Salesforce.
Flint represents an early public test of what Recursive calls recursive self-improvement — the concept that AI can help build better AI, which then builds better AI still. The company’s longer-term goal, according to CEO Richard Socher, is for the Eureka Machine to achieve full recursive self-improvement with human oversight. Whether that cycle can scale meaningfully remains one of the more contested questions in AI research, with some researchers warning such systems could become difficult to control.