Voice is human. Fast. Intuitive. But traditional voice AI has a dirty secret: your convos become training data. Maple Voice fixes this: • Your words stay yours • Audio encrypted before leaving your device • Zero data retention or training use Easy to use: 1. Tap 🎤 in any Maple chat 2. Speak naturally (questions, ideas, requests) 3. Hit "Send" 4. Hear Maple respond in a clear voice (and you can see the text) Behind the scenes: 🔒 Microphone data is encrypted on your device 🗣️ Speech-to-text via Whisper (OpenAI) in secure enclaves 🧠 Your chosen AI model processes the request (DeepSeek, GPT-OSS, Qwen3 Coder, etc.) 🔊 Voice replies via Kokoro text-to-speech (also in enclaves) Available NOW for all Maple Pro, Team, & Max users. Free/Starter? Upgrade to unlock →

Maple AI

Private AI chat

Works on web, desktop & iOS. Proxy users coming soon. Voice shouldn't mean surrendering privacy. With this release, we're proving you can have both: the convenience of spoken AI and the security Maple delivers. Your voice is yours—not a data point to be mined and sold. Read and share the full release article:

Maple AI Blog

Speak Freely – Encrypted Voice Arrives in Maple

What if you could talk to AI as naturally as you talk to a colleague—with the same ironclad privacy Maple delivers? Starting today, you can. W...

Let’s go!!!!

Can web be accessed?

Wow

So is this deployed to AWS using open secret protocol?

Wow, great feature! Will try it out asap.

This is awesome! When should we expect the app update?

Are you guys shipping on android?

@Thomas I’ve made it easy for anyone who wants to connect with me privately to have a quick conversation. Please feel free to reach out to me directly through the SimpleX app using the link below,I’ll be glad to chat with you.

SimpleX Chat - Invitation

is this service for free ? where available ?

Killing it guys

🗣️🤖🔥🔥 View quoted note →

Wow

Hmmm, i would bet gov or its vendors would have every americans' voice print now, so could I be doxxing yourself?

Will surely be trying this beauty..🧡🧡

Does it upload the original audio file or is it a simple voice to text?

You can estimate geographical location pretty close based on the words we choose.

Awesome! What’s your honest opinion on another way so use AI privately, with nostr, LN, cashu and the service @routstr?

Routstr

Routstr - Decentralized AI Inference Router

Access top AI models with Bitcoin Lightning payments. Permissionless, private, and decentralized.

Routstr

The future of AI access is permissionless, private, and decentralized

Yesterday I was just talking about how great this would be. Must’ve heard me talking about it to ChatGPT. 🤣 View quoted note →

Give them a try. We need more freedom tech tools in this space. I like the concept of what they're building and that people will have choices for UX, payment models, etc. Weigh the privacy risks and pick what you're comfortable with.

I’m grateful for the reply, @Marks

@Maple AI Look at the #Android, please...

👀

#Android 1st

Fantastic work!👏 Now that Whisper is integrated, if you added the ability to transcribe long audio files – 1 to 5 hours – it would become the perfect AI tool.

Android app would be appreciated. Don't want to have to give my browser mic permissions.

But in your case, you only retain users data stripped of PII for 48 hours? Are you trying to tell me that you don't train on that data in that 24 hour 48 hour period? This seems highly sus as my kids would say. Where is your decentralization protocol? Where is your verifiable scarcity? Where is your actual proof of all you say? Love the concept! Don't trust verify.

The audio file is encrypted on device, uploaded to Whisper running in the secure enclave, then text is sent back to your device and passed through the normal Maple secure LLM flow