Bodhi App - Run LLMs Locally

Intro

Bodhi App is described as a unified AI gateway that brings together local models, cloud APIs, and MCP tool integration. Documentation and site snippets indicate compatibility with the OpenAI API and components from the HuggingFace ecosystem and llama.cpp. The project ships a built-in chat interface and references enterprise access control and local model deployment options, positioning it as a bridge between on‑premise inference and cloud‑based services.

Key Features

• Unified gateway for local model inference and cloud API proxying (described in official materials).
• Built‑in full‑featured chat UI available at /ui/chat/ for conversation management and tool workflows.
• Integration points with MCP tools enabling tool calling and agent-style workflows in chat.
• Compatibility notes include the OpenAI API, the HuggingFace ecosystem, and components from llama.cpp.
• Installation options mentioned for desktop and Docker, including a guided setup wizard.
• Enterprise access control is referenced; privacy tags appear in source metadata.

The available evidence does not include pricing, detailed security or data retention practices, supported model sizes, performance benchmarks, or exhaustive lists of supported tools and integrations.

Who this is for

This profile is aimed at teams and technical users who need a single interface to combine local inference and cloud APIs, want a built‑in chat UI for conversational workflows, or plan to integrate external tools into agent-style chats. The documented desktop and Docker installation paths with a guided setup suggest it is suitable for users who can manage local deployments or containerized installations. Organizations requiring explicit pricing, security guarantees, or detailed performance specifications should note those items are not present in the provided evidence.