The next era of AI doesn't live in data centers: it lives in your browser tabs.
- Runs on the client: WebGPU when available, CPU otherwise, chosen by what the device supports.
- One Rust source for native and the browser.
- Checked against the reference implementation before any speed work.
Every project on one page: idleintelligence.org · Models and data: huggingface.co/idle-intelligence
See some of these models at work on trucs.ai, and read how they were built on trucs.ai/blog.
