At Aistatine, we’re building on-device AI models that run entirely on your device and help you build production-ready code. No cloud inference and no code sent to a hosted AI service.
Built by the Minds Behind
user@aistatine ~ %
V1 introduces an on-device AI model that runs locally on your hardware, generates production-ready code, and answers complex technical questions. Built to compete with leading AI models, it helps developers build, debug, and ship software faster while keeping data on-device.
Core Capabilities
Your code stays on your machine. No cloud inference, no data sent to a hosted service. See how we compare →
Understand, refactor, and navigate massive codebases with up to 1 million tokens of context.
Built with a security-first architecture. No telemetry, no data collection, complete privacy. Read our privacy policy →
A slash command for every workflow. Build, search, debug, and ship without leaving the terminal.
Model Benchmarks
One view across coding, knowledge, reasoning, and vision benchmarks.
SWE-bench Verified
MMLU
GPQA Diamond
MMMU
Who it’s for
Run @code locally and keep your code out of the cloud — whether you’re shipping solo, working across a team, or operating in a regulated environment.
Pick your platform and run a single command. It installs the @code CLI, platform helpers, and local inference engine, then downloads the model on first run — everything stays on your device.
aistatine@Mac ~ % curl -fsSL https://raw.githubusercontent.com/aslam-anees/at-code/src/install/install.sh | bash@, @code or atcode to start.