gabber  by gabber-dev

AI app builder for real-time multimodal experiences

Created 8 months ago
1,092 stars

Top 34.6% on SourcePulse

GitHubView on GitHub
Project Summary

Gabber is a real-time, multi-modal AI application engine designed for developers to build complex AI workflows involving voice, text, and video. It targets developers seeking a flexible, graph-based approach to AI application development, enabling the creation of interactive experiences with multiple participants and simultaneous media streams.

How It Works

Gabber utilizes a graph-based architecture where applications are composed of interconnected "nodes" that perform specific AI tasks. Data flows between nodes via typed "pads," allowing for real-time processing of various media streams. The system includes an engine for running these graphs, an editor for their creation, and a repository for storage. It also features a state machine system for managing application logic and transitions.

Quick Start & Requirements

  • Install: make all (installs LiveKit, uv, and runs backend services).
  • Prerequisites: Homebrew (for LiveKit), uv (Python dependency manager), Node.js.
  • Secrets: API keys for nodes (e.g., OpenAI) should be placed in a .secret file.
  • Access: Frontend available at http://localhost:3000.
  • Docs: Gabber repo (SDK documentation within).

Highlighted Details

  • Supports real-time, multi-modal AI across voice, text, and video.
  • Graph-based application building with reusable SubGraphs.
  • Integrated state machine for complex application logic.
  • Offers JavaScript/TypeScript and React SDKs, with Python, Unity, iOS, Android, and React Native SDKs planned.

Maintenance & Community

  • Actively updated with new nodes and functionality.
  • Community engagement via GitHub Discussions and Issues.
  • Discord server available for feedback and updates.
  • Follow on Twitter/X: @gabberdev.

Licensing & Compatibility

  • Gabber engine and frontend code are under the Sustainable Use License and Gabber Enterprise License (source-available, self-hostable, extensible).
  • Examples and SDKs are licensed under Apache 2.0.
  • Compatibility for commercial use or closed-source linking depends on the terms of the Sustainable Use License.

Limitations & Caveats

Several SDKs (Python, Unity, iOS, Android, React Native, Flutter) are listed as "Coming Soon," indicating a focus on JavaScript/TypeScript and React at present. The licensing model may have restrictions for certain commercial applications.

Health Check
Last Commit

3 months ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
8 stars in the last 30 days

Explore Similar Projects

Starred by Sebastian Raschka Sebastian Raschka(Author of "Build a Large Language Model (From Scratch)"), Pawel Garbacki Pawel Garbacki(Cofounder of Fireworks AI), and
16 more.

chainlit by Chainlit

0.3%
12k
Framework for rapid conversational AI app development
Created 3 years ago
Updated 2 days ago
Feedback? Help us improve.