Getting started

LiteGate is a local, native AI gateway in development. Its goal is to connect OpenAI- and Anthropic-compatible tools to local and cloud models through a single endpoint.

Current status

There is no usable gateway or public installer yet. The current foundation lets you build the workspace and validate configuration. The daemon does not yet open an HTTP server, and status does not query a running service.

Explore the project

The code is available on GitHub under Apache 2.0. To work on it, you need Git and Rust with rustup; rust-toolchain.toml in the repository pins the compiler version.

From a terminal:

git clone https://github.com/silverbacking/LiteGate.git
cd LiteGate
cargo build --workspace
cargo test --workspace

Validate an example configuration

Run from the repository root:

cargo run -p litegate-cli -- config validate --config tests/fixtures/config/valid-phase1.toml

The expected output is Configuration is valid. This operation checks the file; it does not connect to a provider or send inference requests.

Next steps

• Configuration: format, validation and file selection.
• Clients and providers: current scope and planned connections.
• Installation and updates: package availability.
• Troubleshooting: common issues at this stage.

See the roadmap for later implementation phases.