onyx — What is it?

Onyx is an open-source AI platform that provides a feature-rich interface for interacting with various Large Language Models (LLMs), enabling advanced functionalities like RAG, web search, code execution, and more.

⭐ 29,916 Stars 🍴 4,063 Forks Python Author: onyx-dot-app
Source: README View on GitHub →

Why it matters

Onyx is gaining attention due to its comprehensive feature set that addresses the need for a robust AI interaction platform. It fills the gap by offering a unified interface for various LLMs, including both self-hosted and proprietary models. The project stands out for its advanced capabilities like RAG and deep research, which are unique in the open-source space.

Source: Synthesis of README and project traits

Core Features

Agentic RAG

Onyx implements a hybrid index and AI Agents for information retrieval, providing high-quality search and answer capabilities. It is set to be benchmarked soon.

Source: README
Deep Research

Onyx offers in-depth research reports through a multi-step research flow, ranking top on the leaderboard as of February 2026.

Source: README
Custom Agents

Users can build AI Agents with unique instructions, knowledge, and actions, enhancing the platform's adaptability.

Source: README
Web Search

Onyx supports web browsing for up-to-date information, integrating various search engines and providing in-house web crawling capabilities.

Source: README
Artifacts

The platform can generate documents, graphics, and other downloadable artifacts, expanding its utility beyond text-based interactions.

Source: README
Actions & MCP

Onyx allows agents to interact with external applications, offering flexible authentication options for secure integration.

Source: README
Code Execution

Onyx supports executing code in a sandbox, enabling data analysis, graph rendering, and file modifications.

Source: README
Voice Mode

Users can chat with Onyx via text-to-speech and speech-to-text, enhancing accessibility and user experience.

Source: README
Image Generation

Onyx can generate images based on user prompts, adding a visual dimension to its capabilities.

Source: README

Architecture

The architecture of Onyx suggests a modular design with clear separation of concerns. It features a backend for handling business logic, a model server for AI model inference, and a user interface for interaction. The code tree indicates the use of various design patterns, such as the use of Docker for containerization and the integration of multiple dependencies for different functionalities.

Source: Code tree + dependency files

Project Knowledge Graph

Knowledge graph: project (center) + core features (inner hexagons) + key dependencies (outer chips) aioboto3 cohere fastapi google-genai litellm Agentic RAG Deep Research Custom Agents Web Search Artifacts Actions & MCP Code Execution Voice Mode onyx Project Core feature Key dependency

Center: project; inner ring: core feature modules; outer ring: key dependencies. Auto-generated from core_features and tech_stack.key_deps.

Tech Stack

LanguagePythonFrameworkFastAPI, Prometheus, Sentry, Uvicorn
aioboto3coherefastapigoogle-genailitellmopenaipydanticprometheus_clientsentry-sdkuvicornvoyageai
Docker, Kubernetes, Helm/Terraform
Source: Dependency files + code tree

Quick Start

``` curl -fsSL https://onyx.app/install_onyx.sh | bash ```
Source: README Installation/Quick Start

Use Cases

Onyx is suitable for developers and organizations looking to integrate advanced AI functionalities into their applications. It is useful for scenarios such as building AI-powered chatbots, conducting in-depth research, and automating tasks through custom agents.

Source: README

Strengths & Limitations

Strengths

  • Strength 1: Comprehensive feature set for AI interaction
  • Strength 2: Support for various LLMs
  • Strength 3: Modular and scalable architecture

Limitations

  • Limitation 1: Unknown licensing
  • Limitation 2: Potential complexity in deployment and maintenance
Source: Synthesis of README, code structure and dependencies

Latest Release

v4.0.0-beta.0 (2026-05-20): Fixed issues and added features, including improvements to external group sync and agent selector searchability.

Source: GitHub Releases

Verdict

Onyx is a promising open-source AI platform that offers a wide range of features for developers and organizations. Its comprehensive feature set and support for various LLMs make it a valuable tool for those looking to integrate advanced AI functionalities into their applications. It is particularly suitable for teams that require a robust and scalable AI interaction platform.

Transparency Notice
This page is auto-generated by AI (a large language model) from the following public materials: GitHub README, code tree, dependency files and release notes. Analyzed at: 2026-05-24 13:25. Quality score: 85/100.

Data sources: README, GitHub API, dependency files