Self-host open-assistant-api
OpenAI-compatible self-hosted assistant API with RAG
- Stars
- 369
- Ports
- 8086 (API server), 80 (Playground UI)
- Requires
- MinIO (S3-compatible storage), MySQL, Redis
samepaage/open-assistant-apiWhat is open-assistant-api?
Open Assistant API provides a drop-in replacement for the OpenAI Assistants API that you can run on your own infrastructure. It implements the OpenAI-compatible REST interface so existing code using the official OpenAI Python client works unchanged. The project adds retrieval-augmented generation through the R2R engine, optional web search via the Bing API, and an extensible tool system for custom functions. It supports multimodal inputs and streaming message output. Deployment is via Docker Compose, which spins up the API server, a Celery worker, a playground UI, MySQL, Redis, and MinIO (S3-compatible object storage). Authentication uses Bearer tokens with optional admin-token-protected token management for multi-tenant SaaS-style deployments. The codebase is MIT-licensed and written in Python 3.10 using FastAPI, Celery, SQLModel, and LangChain. There are no published releases yet; the main branch receives regular commits.
Should you self-host open-assistant-api?
open-assistant-api scores 44/100 on selfhost.directory (how scoring works) and ranks #71 of 120 in AI Assistant. It has 369 GitHub stars.
There was no stable release in the past 12 months, so check the repository's recent activity before you rely on it.
Setup is moderate: expect some configuration beyond a single command. An official Docker image is available. It is MIT-licensed and written mostly in Python.
Its closest alternative on selfhost.directory is OpenOpenAI (45/100, 549 stars). It scores higher overall, so compare the two before you pick. See open-assistant-api vs OpenOpenAI side by side.
Features
- OpenAI Assistants API Compatibility
open-assistant-api README
More in LLM Integration
OpenGateLLMSelf-hosted open-source API gateway unifying generative AI models with OpenAI-compatible endpoints, access control, rate limiting, and usage tracking.
OpenKnowledgeOpen-source, self-hosted WYSIWYG markdown editor and LLM wiki with MCP agent editing, git-backed sync and team sharing.
open-notebookSelf-hosted, open-source Notebook LM alternative: research notebooks with cited AI chat, full-text and vector search, transformations, podcasts and 20+ AI providers.
Open ResponsesA self-hosted, open-source server implementing the OpenAI Responses API specification with support for multiple LLM providers and the Agents SDK.
OpenVikingOpenViking gives AI agents one self-hosted, browsable viking:// filesystem for memory, knowledge RAG and skills; open-source (AGPL-3.0), deployed via Docker or pip.
Open-WebUIOpen WebUI is a self-hosted AI platform connecting to Ollama and OpenAI-compatible APIs with RAG, plugins, voice/video calls, and enterprise auth on your hardware.
OpenWorkSelf-hosted and open-source, this Electron workspace puts AI agents to work on local files; over 50 providers, Ollama support, and a self-hosted Den control plane.
oss-llmops-stackAn open-source, self-hosted LLMOps platform pairing LiteLLM's unified model gateway with Langfuse's tracing, evaluation, and prompt tooling for production AI workloads.
PaddleOCROpen-source, self-hosted OCR and document-parsing toolkit: scanned PDFs and photos come out as Markdown or JSON for RAG pipelines, across 100+ languages.
PangolinSelf-hosted, open-source SASE stack combining a WireGuard zero-trust VPN, identity-aware reverse proxy, PAM and an AI gateway; the AGPL Community Edition is free.
parlantOpen-source, self-hosted conversational harness for customer-facing AI agents: guidelines, journeys, canned replies and tools matched per turn by a FastAPI server.
picoclawSelf-hosted, open-source AI assistant in Go: a single binary under 10MB RAM with cron jobs, MCP tools, web search and 19+ chat channels like Telegram and Discord.

0 comments