Media & Creative

Voice

Record a message, transcribe speech, hear your AI talk back.

Talk to MoClaw in voice and get a voice reply. Record a message in any chat surface, MoClaw transcribes it, generates a response, and (if you want) reads the response back in a chosen voice. Works in the web app, the iOS and Android Telegram bot, and any Slack DM.

So funktioniert es

3 Schritte, um Voice einzubinden — ohne Engineering-Aufwand.

  1. 1

    Tap the mic in any chat

    On web, hold space-bar. On Telegram, send a voice note. On Slack, attach an audio file.

  2. 2

    MoClaw transcribes and replies

    Whisper-class transcription with multi-language support. The reply comes back as text by default.

  3. 3

    Or have it read back to you

    Toggle 'voice reply' in Settings to get audio replies in a voice you pick (ElevenLabs, OpenAI TTS, or system voices).

Probiere zu sagen

Echte Prompts, die du in Voice einfügen kannst.

  • Voice memo: 'Remind me to follow up with the design team about the new homepage by EOD Wednesday.'
  • While driving: 'Read me the headlines from Hacker News, top 5.'
  • Send me a daily morning brief as a 90-second voice note instead of text.

Schritt-für-Schritt-Demo

Was tatsächlich passiert, wenn du den Prompt sendest.

Prompt 01 4 Schritte

“Voice memo from the car: 'Email Sarah saying we're a yes on the Q2 partnership and ask her to send the contract.'”

Was MoClaw tut

  1. 1 Transcribes the voice note using Whisper.
  2. 2 Looks up Sarah in your Gmail contacts (most recent thread).
  3. 3 Drafts the email in your voice — short, decisive, asks for the contract.
  4. 4 Asks for confirmation in chat (since it's a real outbound email).
Ergebnis

Reply pops up: 'Drafted to sarah@partner.co. Subject: Q2 partnership — yes from us. Body: Sarah, we're a yes on the Q2 partnership. Can you send the contract this week? Best, [you]. Send?' You tap send while still driving.

FAQ

Kurze Antworten zu Preisen, Datenschutz und Limits.

What languages does transcription support?
100+ via Whisper. Auto-detected per recording. Mixing languages in the same recording works too.
Which voices can I pick?
On free, OpenAI TTS voices (Alloy, Echo, Fable, Onyx, Nova, Shimmer). On paid, the full ElevenLabs catalog plus voice cloning from a 30-second sample.
Does it work hands-free?
On the iOS Telegram client, yes — voice in, voice out, no taps. On the web, you need to start each turn with space-bar (browser security restriction).
Where does my audio go?
Encrypted in flight, transcribed on a hosted Whisper instance, audio deleted within 60 seconds. Transcripts stay in your chat history (encrypted, deletable from Settings).

MoClaw kostenlos testen.

1.000 Credits pro Monat oder eigenen Key mitbringen für unbegrenzte Nutzung.

Jederzeit kündbar