Report Abuse

Basic Information

ChatPilot is a self-hosted chat agent web UI that provides an end-to-end interface for interacting with agent-driven conversational systems. It implements AgentChat dialogs and agent assistant calls based on the Agentica project and is intended to let users run and interact with agents that can call external tools, fetch and parse web content, run code, and answer questions over uploaded files. The project separates frontend and backend, using a Svelte frontend and a FastAPI backend, and includes Docker and local startup instructions. It supports multiple LLM backends including OpenAI and Azure, plus integrations for Ollama and litellm, and provides RAG capabilities for file-based QA. The repository includes build instructions for the web UI and configuration examples for different LLM providers.

Links

Categorization

App Details

Features
The README describes tool calling support, including Google search via Serper or DuckDuckGo, an automatic URL parsing tool that reproduces Kimi Chat style website parsing, and a Python code interpreter that can run in E2B virtual environments or local Python. It supports multiple LLM connection methods such as OpenAI/Azure API, Ollama for local open models, and litellm for cloud-deployed models. RAG for file question answering is supported. The project is architected with a Svelte frontend and FastAPI backend, offers Docker images for quick deployment, supports voice input and output, image generation, and includes user management, permission control, and chat import/export features. Frontend build artifacts or a build.zip are provided and npm build instructions are included.
Use Cases
ChatPilot is useful for teams or individuals who want a ready-made web interface to run agent workflows and experiment with tool-augmented LLM interactions without building a UI from scratch. It lets users connect different LLM providers and local models, enabling comparison or fallback strategies. Built-in tools such as web search, URL parsing, and an executable Python interpreter make it practical for tasks that require browsing, fetching web content, or running code during a conversation. RAG support enables document and file question answering workflows. Docker and local install instructions simplify deployment, while frontend/backend separation allows customization of the UI or backend logic. User management and export features help with evaluation, demos, or multi-user setups.

Please fill the required fields*