Ed

Embed Ed. Your app gets a brain. Your users keep their data.

Ed is the agent layer above Onde. It loads an open-weights model, keeps the conversation, calls only the host tools you register and hands back replies. The model runs on the user's device, so there is no metered inference API and private context stays put. The ed-acp crates build ACP agents (Onde Code, SplitFire Agent, siGit) on the same base.

Ed doesn't own sessions, persistence, model catalogues or downloads. Those differ per application and stay yours.

Install

cargo add ed-agent   # embedded chat
cargo add ed-acp     # ACP agent

For Swift, add the package from ondeinference/ed-swift.

Quick start

use ed_agent::{Ed, GgufModelConfig};

let ed = Ed::new();
ed.load(GgufModelConfig::qwen25_1_5b(), None, None).await?;

let reply = ed.send("Summarise this thread.").await;
println!("{}", reply.reply.unwrap_or_default());

Streaming, history, tools and approvals are covered in crates/ed-agent.

Use it your way

• In your app: ed-agent, plus ed-agent-tauri for Tauri and ed-agent-ffi for Swift.
• As an ACP agent: ed-acp with a Profile, and ed-acp-tui for a terminal UI.
• Inside a Tauri app: ed-acp-tauri runs an ACP agent in process or an external stdio agent.
• With Claude: ed-claude, a profile on the Anthropic API with the user's key.
• With MCP servers: ed-mcp. Onde accounts: ed-onde-account. Tests: ed-acp-testkit.

In production

smbCloud MailX, Siti AI (the open reference app), Pepak Basa Jawa Komplit and Rumi.

Docs

Each crate has its own README: ed-agent, ed-agent-tauri, ed-agent-ffi, ed-acp, ed-acp-tui, ed-acp-tauri, ed-claude, ed-mcp, ed-onde-account, ed-acp-testkit.

Contributing

Open pull requests against development. Before you push:

cargo fmt --all -- --check && cargo clippy --locked --workspace --all-targets -- -D warnings && cargo test --locked --workspace

License

MIT or Apache-2.0, at your option. See LICENSE-MIT and LICENSE-APACHE.

Copyright

Free the intelligence.

© 2026 Onde Inference.