Lucy TTS is a small Mac and iPhone app for people who would rather type than speak in the moment. You type, press Enter, and Lucy speaks the line out loud in a voice you chose. While Lucy is speaking, you can keep typing. The next line goes in a queue and plays in order. Nothing gets lost.

Lucy uses Fish Audio for the actual voice generation. You bring your own Fish Audio account and API key. Lucy doesn’t have its own servers; nothing you type ever goes through us.

Beta status. Lucy is in private beta as of June 2026. The macOS app is usable. The iPhone app is invite-only via TestFlight. Open issues are tracked on GitHub.

Start here

1. Setup guide → — what you need, what it costs, and how to install Lucy step by step.
2. Privacy → — exactly what leaves your device and what doesn’t.
3. Troubleshooting → — common snags, in plain English.

Who Lucy is for

• People who can type comfortably but find speaking out loud tiring, painful, slow, or socially difficult — temporarily or long-term.
• People who use Lucy to participate in video calls, in-person meetings, or one-on-one conversations.
• People who already have (or are willing to create) a Fish Audio account and pay Fish’s per-character usage fees.
• People who are comfortable following kit-style instructions with screenshots.

Who Lucy (currently) isn’t for

• People who need a clinical AAC (augmentative and alternative communication) device. Lucy is a communication aid, not certified medical equipment.
• People who rely on a screen reader to use their device. Lucy does not include VoiceOver/screen-reader integration and that is not on the roadmap. If that’s what you need, please look at established AAC apps designed around it.
• People who want a single all-in-one subscription. Lucy is BYOK (bring your own key) and you pay Fish Audio directly.
• People who are not comfortable installing apps outside the App Store. The macOS build is currently distributed as a signed .app you download yourself. The iPhone build is distributed through Apple’s TestFlight, which is also outside the regular App Store.

What Lucy costs

Lucy itself is free and the source code is open under AGPL-3.0. You don’t pay us.

You do pay Fish Audio for the actual voice generation, on their own pricing. As a rough sanity check, a typical conversation sentence is around 80–150 characters; Fish typically bills well under a cent per sentence at current rates. A reasonable starting budget for testing is about $5 of Fish credit. Always check Fish’s pricing page for the current rate.

To avoid surprise bills, keep your Fish account balance low and top it up as needed. A Lucy-side daily character cap is planned but not yet built.

What’s on the roadmap

These are tracked as GitHub issues so you can follow progress:

• Cancel TTS in flight when Stop is pressed during generation — today, Stop only halts playback; a cancelled sentence may still play.
• Daily character cap and in-app usage display — so a stuck loop or a long paste can’t spike your Fish bill.
• Persistent on-device speech history — today, history is cleared when Lucy quits.
• In-app editing of phrase presets — today, phrases are read-only.

Want to suggest something?

Open a GitHub issue with the feature request label. We read every one.

For security or private matters, see the Contact section below.

Contact

• Bug reports and feature requests: GitHub Issues — please tag with bug or feature request.
• Security disclosures and private messages: lucychapar@pm.me.

Honest disclaimers

• Lucy is built and maintained by one person with AI assistance. Response times will vary. We will not silently abandon the project, but we also will not pretend to be a larger team.
• Lucy is not affiliated with, sponsored by, or endorsed by Fish Audio. Fish Audio is a separate company with their own Terms and Privacy Policy, and you agree to those directly when you sign up with them.
• “Lucy TTS” and the Lucy logo are reserved marks; see TRADEMARKS.md.