VersAIt
/ver · say · it/

Just say it.

Voice-native AI coordination. Talk to Claude, ChatGPT, Gemini, and your own agents — together, hands-free, with each agent speaking in their own voice. Your private agent takes the minutes. The cloud AIs do the heavy thinking.

Join the TestFlight How it works →

The meeting place for you, your AI, and your AI's memory.

Cloud AIs are powerful but forget. Your home agent remembers everything but can't reason like Claude. VersAIt is where they sit down together and work as a team — through voice, through text, on your terms.

🎙️

Voice-first

Talk naturally. Dictation is additive — pauses don't erase your thought. Replies come back as TTS in distinct voices per agent, so you can tell who said what without looking.

👥

Multi-agent group chat

Claude, ChatGPT, your local agents — any mix in a single conversation. Independent mode for second opinions. Shared context for discussion. Pipeline for draft-then-review.

🔐

Your agent is yours

Local agents run on your PC. Your memory, your files, your privacy. The cloud AIs never see what's private unless your agent specifically surfaces it during the conversation.

"Claude is powerful. Your agent remembers. VersAIt is where they meet." — The VersAIt product vision, distilled

How it works

Four-minute setup. Then hands-free use forever.

Install VersAIt

Coming to the App Store. TestFlight beta opens before launch — join the waitlist below.

Connect a cloud AI

Bring your own Anthropic / OpenAI / Google API key. One-tap quick-connect buttons. Your key stays on device.

Run a home agent (optional)

Install a VRP1-compatible memory agent on your PC. Or start with cloud-only and add one later.

Talk

Hit the mic, speak, pause. The agents reply aloud in their own voices. Swipe for text mode.

Works with

Claude (Anthropic) ChatGPT (OpenAI) Gemini (Google) Any VRP1-compliant local agent Hermes OpenClaw Your own — bring a spec, we'll accept it

Questions

How is this different from ChatGPT's voice mode?

ChatGPT's voice mode is one app talking to one model, with no persistent memory between sessions. VersAIt lets you bring Claude, ChatGPT, Gemini, and your own private memory agent into a single conversation — your agent keeps the history, the cloud AIs do the reasoning. It's a team, not a chatbot.

How is this different from Wispr Flow?

Wispr Flow is premium dictation — speech-to-text cleanup for writing. VersAIt also includes text-to-speech for every reply and multi-agent orchestration. Wispr helps you write; VersAIt is a two-way conversation with multiple AIs.

Do I need a home agent to use VersAIt?

No. You can use VersAIt purely with cloud AIs. A home agent adds persistent memory and privacy — it's where the product gets magical — but it's optional.

What does it cost?

TestFlight beta is free. Pricing TBD for the App Store launch; expect a modest one-time purchase or an optional subscription for cloud-side conveniences. API usage for Claude/ChatGPT/Gemini is billed directly to you by the provider — VersAIt never marks those up.

Is my data private?

Every API key is stored in the iOS Keychain. Speech recognition runs on-device (Apple's SFSpeechRecognizer). Your conversations go to whichever provider you configured, nothing else. We don't log, sell, or analyze your conversations.

What does "VersAIt" mean?

Pronounced "ver-SAY-it." It's a small pun — just say it — and "AI" sits inside the name to mark what it's about. The logo makes "AI" prominent in red.

Is VRP1 open?

Yes. VRP1 (VersAIt Runtime Protocol) is an open specification for any HTTPS-speaking backend. If you want to build an agent that works with VersAIt, implement VRP1 and you're in.

Coming soon.

VersAIt is in private testing today. Join the TestFlight waitlist to try it before the public launch.

Join TestFlight waitlist