How to start:

• Provide your Google AI API key (see processing notes below)

• Upload a file

• Viola! Your transcription appears super fast (using response streaming) 🥳

Features:

• Lightning fast transcription speed ⚡

• 1500 free transcriptions per day 😱

• AI custom instructions (example: “Don't include 'um' or 'uh'”)

• The transcriber can handle files with up to 8 hours of audio 🔥

• Open source on GitHub

How does it work?

This free tool takes advantage of the multimodal capabilities of Gemini 2.0 Flash. Gemini 2.0 Flash can handle both text and audio input! Your API key is used to send requests to the Google API. It is stored locally in your browser and is never sent to our servers.

Notes:

• If you have a really long audio file, you might not get a full response, since the Google API limits the amount of tokens a model can generate in a single response.

• MP3 files are recommended, but some other formats are supported.

• Files larger than 25 MB will be compressed to fit within the Google API limitations. If your file is still 25 MB after compression, you can't submit it 🥹

• The site does not use cookies 🍪

API Key Processing Notes

When you provide your API key to the tool, it is saved locally in your browser (using local storage). It is used only to authenticate your API requests to Google; it is never sent to our servers, collected, or tracked in any way.

You can remove your API key at any time by clicking the 'Reset API Key' button in the top right of the tool view, removing the key, and clicking 'Save'.

If you are still uncertain, feel free to review the API key handling logic in the source code here:

https://github.com/The-Best-Codes/audio-transcription-gemini/blob/main/app/components/TranscriptionApp.tsx#L252

Happy Transcribing!