An open-source TypeScript platform for turning open-weight models into downloadable AI apps with built-in inference and a multi-agent runtime.
Lloyal homepage hero
1 / 2Overview
What is Lloyal?
Ship apps that think and act.
Lloyal is an open-source platform from LLoyal Labs for building AI apps on open-weight models. You build your 'harness' in TypeScript starting from a working scaffold (npx lloyal-ai new), with built-in inference and a multi-agent runtime — no OpenAI/Anthropic key, no Docker, Ollama, LangGraph, or vector database required.
Its 'Abilities' plugin system is inference-native: extensions run inside the model's execution, can see the caller, share live inference state, and fork new agents. One signed channel hosts first- and third-party Abilities that install with a single command.
The same codebase ships as a terminal/TUI app, a desktop app (Electron/Tauri), or a web app; mobile (iOS/Android) is marked as coming soon. Apps can run as a single on-device binary offline, as a multi-user service on a private appliance, or on frontier compute. LLoyal Labs is an NVIDIA Inception program member.
Lloyal does not publish pricing; documentation notes local runs carry no per-token bill. Commercial contact goes through a 'Get in touch' form for partnerships and hardware/OEM.
Platforms and languages
Lloyal Availability
Platforms
Languages
Capabilities
Lloyal Key Features
Vertical Inference architecture
App, model runtime, and live inference state become one programmable system; multiple agents share the resident model's prefix KV cache.
Inference-native Abilities
Plugins run inside model execution rather than behind a network boundary; a signed channel of first- and third-party Abilities installs with one command.
One-command scaffold
`npx lloyal-ai new` generates a working TypeScript AI app with built-in inference, a resident model, and a multi-agent runtime.
Multi-target shipping
Ship the same codebase as a terminal/TUI app, desktop app (Electron/Tauri), or web app — offline on-device, on a private appliance, or on frontier compute.
Best for
Who uses Lloyal?
Developers building downloadable AI apps on open-weight models
Scaffold a TypeScript AI app with built-in inference and a multi-agent runtime, then ship it to terminal, desktop, or web
Common questions
Lloyal FAQs
What is Lloyal?
An open-source TypeScript platform for building AI apps on open-weight models. You scaffold a working app with `npx lloyal-ai new` and ship it as a terminal, desktop, or web app with the model bundled in.
Do I need an OpenAI or Anthropic API key?
No. The model runs inside your app — Lloyal advertises no external API key, Docker, Ollama, or vector database as required.
How much does Lloyal cost?
Pricing is not published on the official site, so it cannot be verified. The docs state that running locally carries no per-token bill.
Which platforms can my Lloyal app target?
Terminal/TUI, desktop (Electron/Tauri), and web browsers today; iOS and Android support is marked as coming soon.
Where is the documentation?
Developer docs live at docs.lloyal.ai, covering setup, Abilities, and deployment.
Reviews
See what the community thinks and share your experience.
—
Based on 0 ratings
Rating distribution
Community reviews
No reviews
No written reviews yet.
