What is better-chatbot?
better-chatbot brings AI conversations, agents and workflow automation into a self-hostable workspace. Users can attach files, assign agents instructions and tools, and group conversations and documents into projects. Project documents are indexed and searched using Postgres with pgvector, while reusable skills provide capabilities that agents load on demand.
The platform can be deployed through Vercel, Docker Compose or a direct installation. Models are accessed through Vercel AI Gateway by default, with alternatives including Ollama, LM Studio and OpenAI-compatible endpoints. An Electron desktop application adds local MCP servers, filesystem access and system keychain support.
The repository describes better-chatbot as the MIT-licensed community edition of keinsaas Navigator. Team administration, per-member AI budgets, organization-wide model policies and other hosted-edition features are separate. Several capabilities, including file uploads, web search and connected applications, require additional service configuration.
Key capabilities
- AI conversations Streaming chat supports tool calls, file attachments, image generation and model selection per message, subject to service configuration.
- Agents and skills Configure agents with roles, instructions and tools, and provide reusable capabilities that load on demand.
- Project retrieval Organize chats, documents and tools in projects, with document indexing and search through pgvector.
- Visual workflows Connect model calls, conditions, HTTP requests, code and templates, then publish workflows as tools available in chat.
- MCP connections Connect tool servers over HTTP or stdio, with OAuth support and per-tool customization; stdio execution requires a self-hosted instance or the desktop app.
- Scheduled execution Run agents on a schedule using configured background services or an authenticated cron setup.
Where better-chatbot fits
better-chatbot fits personal or small-group AI workspaces where conversations need access to project documents, external tools and repeatable workflows. It can also support local-model experimentation or agent-driven tasks on infrastructure managed by the operator.
Before deployment, operators should verify model access, Postgres support for pgvector and the services required by their chosen features. File uploads require Supabase configuration, while scheduling requires configured background services or an external cron setup. Teams should also check whether the community edition meets their administration and budget-control requirements.
Install
docker compose -f docker/compose.yml up -d --buildReviews of better-chatbot
Be the first to share your experience with better-chatbot.
Write a review
Loading review form…
Alternatives to better-chatbot
LibreChatLibreChat brings multiple AI providers, custom agents, file interactions, and conversation management into a self-hosted web interface with multi-user access controls.OOnyxOnyx is an AI platform that indexes connected applications for search, chat and agents, with self-hosting, configurable models and permission-aware retrieval.