speak-gpt  by AndraxDev

Voice assistant based on OpenAI ChatGPT

created 2 years ago
408 stars

Top 72.5% on sourcepulse

GitHubView on GitHub
Project Summary

SpeakGPT is an open-source AI assistant designed for users who want a versatile interface to interact with various large language models and AI services. It supports text generation, image creation, and multimodal input, aiming to provide a comprehensive and customizable AI interaction experience.

How It Works

SpeakGPT leverages a flexible API integration architecture, allowing it to connect with multiple LLM providers including OpenAI, GROQ, Azure, and OpenRouter. This approach enables users to access a wide range of models like GPT, LLAMA, MIXTRAL, GEMMA, and Gemini, offering choice and avoiding vendor lock-in. It supports features like voice input via Whisper and Google, image generation with DALL-E, and GPT-4 Vision for image analysis.

Quick Start & Requirements

  • Install: Download from the SpeakGPT Web link: https://assistant.teslasoft.org/
  • Prerequisites: API keys for supported services (OpenAI, OpenRouter, etc.).
  • Setup: Requires configuration of API endpoints and keys in settings.

Highlighted Details

  • Supports a broad spectrum of LLMs including OpenAI, LLAMA, MIXTRAL, GEMMA, and Gemini.
  • Features voice input (Whisper, Google), image generation (DALL-E), and multimodal capabilities (GPT-4 Vision).
  • Offers extensive customization for model parameters (temperature, topP, etc.) and API providers.
  • Includes a prompt store, chat history management, and an adaptive design.

Maintenance & Community

The project is part of a Bachelor Thesis by Dmytro Ostapenko. Attribution is required for use. Users are encouraged to report bugs and request features via GitHub Issues.

Licensing & Compatibility

Licensed under the Apache License, Version 2.0. This license permits commercial use and linking with closed-source projects, provided the terms of the license are followed.

Limitations & Caveats

The project is noted as part of a Bachelor Thesis, and attribution is required. Some planned features like device routines and chat history sync are not yet implemented. OpenRouter support is limited to text generation.

Health Check
Last commit

1 month ago

Responsiveness

1 day

Pull Requests (30d)
0
Issues (30d)
0
Star History
25 stars in the last 90 days

Explore Similar Projects

Starred by Addy Osmani Addy Osmani(Engineering Leader on Google Chrome), Victor Taelin Victor Taelin(Author of Bend, Kind, HVM), and
1 more.

chatbox by chatboxai

0.3%
36k
Desktop client app for AI models/LLMs
created 2 years ago
updated 6 days ago
Feedback? Help us improve.