Prince Canuma announced that Neywa Labs is coming out of stealth: "We built the infrastructure to build and deploy AI that runs on the devices people already own. Private, offline, unlimited." The company's one mission, as stated on its site, neywalabs.ai, is to make AI available to everyone in the world.
The pitch
Neywa's case is about cost and access. Today, it says, a single AI agent can cost tens of thousands of dollars a month to run, which puts AI out of reach for most builders and for the billions of people who still have no access at all. Local AI changes that: the model runs on the user's own device and the data never leaves it. Neywa calls this "the era of sovereign AI."
Who is behind it
Canuma is the founder and CEO. They are known for the open-source MLX ecosystem for Apple Silicon, with more than 4 million combined downloads across the mlx-vlm, mlx-audio and mlx-embeddings projects, and Neywa is building the inference stack for Apple Silicon. According to coverage of the launch, it lets users run open models, power AI agents and connect developer tools privately. Canuma says Neywa is starting with developers because they set the standard, and its page mentions work with design partners. No funding has been announced.
It joins a growing push toward on-device models, including Liquid AI's LFM3 family and Recursive's Flint local coding model.