AI startup Recursive released its first model, Flint, on Wednesday. Flint is a free coding model that can run on a laptop, and the company says it was built largely with the help of AI.
Recursive’s automated research system, called the Eureka Machine, helped with nearly every step of development. The company says this let a small team build the model from scratch in under four months.
Flint is an early public test of recursive self-improvement, the idea that AI can build better AI, which then builds even better AI. Some researchers worry that AI improving itself could slip out of human control.
For now, Recursive’s version is more limited. It is an AI research assistant overseen by people.
How Flint Works
Because Flint runs on a user’s own computer, developers can use it offline and keep their code private. They can also avoid the usage fees that cloud AI services charge.
The model is free to download on Hugging Face under the Apache 2.0 license. This open source license allows free use and changes.
Flint is a mixture-of-experts model. It uses only 2.5 billion of its 40 billion parameters at a time, which makes it light enough for an ordinary laptop.
By comparison, Meta’s Muse Spark Open, released last week for Macs, uses 816 billion parameters. It needs at least a terabyte of memory.
In Recursive’s tests, Flint produced about 131 tokens per second on a MacBook Pro with Apple’s M4 Pro chip. Similar models, including Google’s Gemma 4 and Nvidia’s Nemotron 3.5 Lightning, managed 58 to 69. On Nvidia’s RTX Spark, Flint topped 150 tokens per second.
On SWE-Bench Verified, a test of how well AI fixes real software bugs, Flint beat other models under 50 billion parameters. Across Recursive’s own average of six tests, only Alibaba’s Qwen3.8 27B scored higher.
Recursive said Flint needs more guidance than large paid coding assistants. It suggested breaking big tasks into smaller steps.
Early reviews were positive. Wired’s Will Knight used Flint on a Mac with 48GB of memory to build small programs, including a to-do list app. He said Flint handled his coding requests every time.
The Role of the Eureka Machine
The Eureka Machine comes up with ideas, runs experiments and uses the results to decide what to try next. According to the company’s launch video, it helped with pretraining, post-training and inference.
Recursive said Eureka’s work includes a sparser model shape and custom kernels that speed up computing. It also developed low-bit training that shrank Flint by 71% with almost no loss in quality.
The company did not say how much time the system saved. OpenAI said last month that by mid-August, its AI agents were logging 3.1 workdays for every one worked by a human researcher.
Recursive came out of stealth in May with $650 million in funding at a $4.65 billion valuation. Backers include GV, Nvidia and AMD Ventures, and CEO Richard Socher is Salesforce’s former chief scientist.
In June, Recursive signed a $410 million compute deal with Amazon Web Services. The company told TechCrunch the deal was “less about headcount and more about agent count.”
Socher told Wired the long-term goal is for the Eureka Machine to reach full recursive self-improvement. He said humans would oversee the process and be able to stop it if something goes wrong.
