What is SpeechGPT used for?

SpeechGPT is a web app for speaking with ChatGPT, with speech recognition and synthesis in over 100 languages, useful for language practice. It is a good fit for practicing speaking a foreign language, hands-free chatting with an AI assistant and experimenting with voice interfaces for LLMs.

Is SpeechGPT open source?

Yes. SpeechGPT is open source under the MIT licence. Its source code is on GitHub at hahahumble/speechgpt and is written mainly in TypeScript.

Yes. SpeechGPT is open source, so the software itself is free to use. A managed cloud version is also available.

Can I self-host SpeechGPT?

Yes. SpeechGPT can be self-hosted on your own server or infrastructure.

What is SpeechGPT an alternative to?

SpeechGPT is an open-source alternative to ChatGPT. Other open-source alternatives to ChatGPT include Chat with GPT, ChatGPT Web and Chatbox.

Is SpeechGPT actively maintained?

Yes. The most recent commit to SpeechGPT was on 11 July 2026, and the latest release is v0.5.1, published on 15 May 2023. The project has 2.7k stars on GitHub.

SpeechGPT: open-source alternative to ChatGPT

About SpeechGPT

SpeechGPT is a web application that lets you converse with ChatGPT by voice. It is aimed at people who want to practice speaking a foreign language or who simply enjoy talking to an AI, and it keeps the experience simple with a chat interface that listens and speaks.

It supports more than 100 languages, includes built-in browser speech recognition and synthesis, and can integrate Azure Speech Services for recognition and synthesis or Amazon Polly for synthesis. The app is mobile friendly and stores all data locally, according to the project. Setup means adding your own OpenAI API key in settings, plus optional Azure or AWS credentials.

SpeechGPT is written in TypeScript and licensed under MIT. It can be deployed on your own hosting, for example with Vercel, and a development guide and changelog are provided, while a public website is also available. It requires paid access to the OpenAI API and optionally to speech services, which the project does not provide. It suits language learners and tinkerers who want a voice interface for ChatGPT.

Key features

Voice conversation with ChatGPT
Speech recognition and synthesis
Support for over 100 languages
Azure Speech and Amazon Polly integration
Local data storage
Mobile friendly interface

Good fit for

→Practicing speaking a foreign language
→Hands-free chatting with an AI assistant
→Experimenting with voice interfaces for LLMs

Built with: TypeScript
Tags: chatgpt; voice-chat; speech-recognition; text-to-speech; language-learning; typescript; openai; web-app

SpeechGPT: questions and answers

What is SpeechGPT used for?: SpeechGPT is a web app for speaking with ChatGPT, with speech recognition and synthesis in over 100 languages, useful for language practice. It is a good fit for practicing speaking a foreign language, hands-free chatting with an AI assistant and experimenting with voice interfaces for LLMs.
Is SpeechGPT open source?: Yes. SpeechGPT is open source under the MIT licence. Its source code is on GitHub at hahahumble/speechgpt and is written mainly in TypeScript.
Is SpeechGPT free?: Yes. SpeechGPT is open source, so the software itself is free to use. A managed cloud version is also available.
Can I self-host SpeechGPT?: Yes. SpeechGPT can be self-hosted on your own server or infrastructure.
What is SpeechGPT an alternative to?: SpeechGPT is an open-source alternative to ChatGPT. Other open-source alternatives to ChatGPT include Chat with GPT, ChatGPT Web and Chatbox.
Is SpeechGPT actively maintained?: Yes. The most recent commit to SpeechGPT was on 11 July 2026, and the latest release is v0.5.1, published on 15 May 2023. The project has 2.7k stars on GitHub.