Using voice input

Speak your prompts instead of typing. Anuma transcribes your voice on your device — your audio is never sent to a server.

Steps

Step 1: Tap the voice button

In any chat, tap the voice input button (the waveform icon) in the chat composer to start recording. If you haven’t enabled voice input yet, tapping it offers to enable it for you.

Step 2: Speak your prompt

Speak naturally. Your words are transcribed in real time as you talk. Tap Stop when you’re done.

Step 3: Review and send

Check the transcription, edit it if you like, and send. Anuma processes it like any typed prompt.

Voice input settings

Voice input lives in Settings → Accessibility (it’s the first section):

Enable voice input — turn the feature on or off
Transcription language — auto-detect works for most speech; pin a specific language if short clips get mistranscribed (this option appears once voice input is enabled)

Good to know

Everything stays on your device. On the web, transcription runs on a local speech model in your browser — the first time you use it, Anuma downloads the model once (roughly 250 MB) and caches it for instant starts after that. No audio is uploaded.
The mobile app uses your device’s built-in speech recognition.
Voice input is available on every plan.
Spoken responses (text-to-speech) aren’t part of Anuma yet — watch What’s new for voice features as they ship.

Related articles