Stories is a free and open source speech-to-text app that turns your voice into text anywhere on your Mac.
Your thoughts flow naturally when you speak them. AI prompts, design ideas, technical reasoning, creative sparks, and personal notes — they all take shape when using your voice.
Stories captures that flow and turns it into text that works everywhere — ready for you and your AI to build on together.
View on GitHub
Whether you vibe-code, vibe-design, or vibe-prototype — your flow begins with voice. Stories keeps that rhythm alive, turning spoken ideas into text your AI can follow and build on.
Talk with your digital friends. Reflect, imagine, or laugh — it all flows easier when spoken. Stories turns shared moments into words your AI can understand, remember, and grow from.
Speak before you write. Ideas, outlines, or rough drafts — say them first, refine them later. Stories captures your thinking so your AI can help turn it into clear, finished text.
Describe what's happening around you — what you see, what's on screen, or what you're working through. Stories gives your AI the context it needs to assist with clarity.
As makers and designers, we needed a voice tool that simply worked — reliable, universal, private, and open. Other transcription app options felt fragile or locked away.
So we built Stories. We use it every day, and now it's free and open source for everyone.
Consistent transcription powered by OpenAI Whisper. Works the first time, every time.
Your audio is sent to OpenAI securely for transcription and deleted right after.
Works anywhere on your Mac. Your words show up right where you're working.
Record and transcribe. Auto or manual paste. No setup, no learning curve — just flow.
Stories uses OpenAI's speech recognition model for transcription. You'll need your own API key (costs ~$0.006 per minute). Get your API key here. Getting your own OpenAI key is easy, follow the steps for this tutorial.
Stories is completely free — we don't charge anything. You just connect your own OpenAI API key and pay them directly for transcription, $0.006 per minute. That's roughly $10 for 30 hours of audio, or around $2 a month for about 50,000 words of transcription — average of a heavy user. If you'd like to support us, you can always donate here.
Stories works with 50+ languages including English, Spanish, French, German, and many more. It automatically detects your spoken language.
Your audio and transcriptions stay entirely on your computer. We never see, store, or access your data. The only external connection is directly from your computer to OpenAI's Whisper API for transcription. You can choose to keep your audio files locally for backup or have them automatically deleted — full control and privacy, always.
Stories includes a basic dictionary feature where you can add custom words and phrases. For example, if you frequently say "Pixelspace" or other technical terms, you can add them so they're transcribed correctly every time. This helps with accuracy for specialized vocabulary.
Stories automatically saves your audio locally and retries failed uploads with smart fallbacks. If something goes wrong during transcription, the app will try again multiple times before giving up. Your recordings are backed up on your Mac (optional) until they're successfully transcribed, so you never lose your work.
Stories is free and open source software under the MIT License. You can view the source code, audit our privacy practices, contribute improvements, or build your own version. We believe in transparency and community-driven development.
Free download. Your OpenAI API key. Unlimited possibilities.