OpenAI Whisper Microservice is a free open-source project on GitHub that you can use to transcribe audio recordings using AI. It only takes your OpenAI API key to transcribe your recordings to text. It does that very quickly and you just have to deploy it locally or on a cloud. In addition, there is a hosted version of this service is available and you can use that if you need a quick speech to text solution.
For now, it can only transcribe spoken words. But I hope in the later update, they add an option to transcribe an audio file after uploading it. It is a good tool for dictation. It supports multiple languages as it uses Whisper underneath. But since Whisper that it uses is from OpenAI itself over API, so you don’t need a PC with high hardware.
How to Convert Speech to Text using this AI based Tool?
This OpenAI Whisper Microservice is fully open source, so that you can grab its source code and host it in a few clicks on a VPS or run it locally. But the developer of this tool is also offering its online hosted version that you can try with your own API key. The main UI is very simple and easy to use.
So, enter your key and then hit the record button. It will ask your permission to access the microphone. So, allow that and then start speaking. If you want to read something that you’d like in the text form, then you can now do that.
After you are done recording, you just need to submit the recording and then wait for a few seconds. It will show you the transcribed voice recording to you. Just copy the text and then use it anywhere you like.
In this way, you can use this simple and open source microservices to transcribe your voice recordings. It is perfect for speech to text and online dictation. You now don’t have to pay hefty subscriptions for other services out there.
Closing thoughts:
We have covered how to use Whisper itself on tour PC but that requires technical knowledge and a powerful hardware to run the AI model. Now, you can automate all that using the free and open-source services. Since, all the heavy lifting is done by OpenAPI API, you can run it on an average PC or laptop.
