Website • [中文]

🌟 Introduction

SpeechGPT is a web application that enables you to converse with ChatGPT.
You can utilize this app to improve your language speaking skills or simply have fun chatting with ChatGPT.

🚀 Features

• 📖 Open source and free: Anyone can use, modify it without cost.
• 🔒 Privacy First: All data is stored locally.
• 📱 Mobile friendly: Designed to be accessible and usable on mobile devices.
• 📚 Support for multiple languages: Supports over 100 languages.
• 🎙 Speech Recognition: Includes both built-in speech recognition and integration with Azure Speech Services.
• 🔊 Speech Synthesis: Includes built-in speech synthesis, as well as integration with Amazon Polly and Azure Speech Services.

📸 Screenshots

📖 Tutorial

1. Set the OpenAI API Key
   
   • Go to Settings and navigate to the Chat section.
   • Set the OpenAI API Key.
   • If you don't have an OpenAI API Key, follow this tutorial on how to get an OpenAI API Key.
2. Set up Azure Speech Services (optional)
   • Go to Settings and navigate to the Synthesis section.
   • Change the Speech Synthesis Service to Azure TTS.
   • Set the Azure Region and Azure Access Key.
3. Set up Amazon Polly (optional)
   • Go to Settings and navigate to the Synthesis section.
   • Change the Speech Synthesis Service to Amazon Polly.
   • Set the AWS Region, AWS Access Key ID, and Secret Access Key (the Access Key should have the AmazonPollyFullAccess policy).
   • If you don't have an AWS Access Key, follow this tutorial on how to create an IAM user in AWS.

💻 Development Guide and Changelog

• For more information on setting up your development environment, please see our Development Guide.
• To view the project's history of notable changes, please check the Changelog.

🚢 Deployment

Deploying with Vercel

Deploying with Docker

1. Pull the Docker image:arm64.

   docker pull hahahumble/speechgpt

2. Run the Docker container.

   docker run -d -p 8080:8080 --name speechgpt hahahumble/speechgpt

3. Visit http://localhost:8080/ to access the application.

Building and running the Docker image

1. Build the Docker image.

   docker build -t speechgpt:arm64 -f Dockerfile .

2. Run the Docker container.

   docker run -d -p 8080:8080 --name=speechgpt speechgpt

3. Visit http://localhost:8080/ to access the application.

📄 License

This project is licensed under the terms of the MIT license.