Tresha AI Assistant · updated 2026-08-08

    How to talk to Tresha with your voice in Dumpster Controls

    Tresha is a two-way voice assistant: tap the microphone, speak your question, and she answers on screen and out loud.

    Before you start

    • A browser that supports audio recording (a current Chrome or Edge is a safe choice). Without support, the microphone button does not appear.
    • Microphone permission for the app when the browser asks.
    1. 1

      Tap the microphone

      The round microphone button sits to the left of the input box, with the tooltip "Tap to speak". Tap it and allow microphone access if your browser asks.

    2. 2

      Speak your question

      While recording, the button glows and animates with your voice, and the input shows "Listening…". Ask naturally, in any of Tresha's languages.

    3. 3

      Stop talking

      Tap the button again ("Tap to stop") or simply pause: after a couple of seconds of silence the recording ends on its own. The input shows "Transcribing…" while your words are converted to text.

    4. 4

      Hear the answer

      Your transcribed question appears in the chat, and Tresha answers both in text and spoken out loud, in the language you spoke.

    5. 5

      Interrupt when you want

      While Tresha is speaking, the input shows "Tresha is speaking: tap to interrupt". Tap the microphone to cut her off and ask the next question, or click the pulsing speaker button in the header ("Stop narration") to just silence her.

    What happens next

    • Only questions asked by voice get spoken answers. Questions you type or send from a suggestion button are answered in text only.
    • Closing the chat panel stops the microphone and any narration immediately. The microphone is never left open in the background.

    Troubleshooting

    I see "Microphone access is blocked. Enable it in Settings to talk to Tresha."

    Your browser or device denied microphone permission. Allow microphone access for the app in your browser or system settings and try again.

    I see "Sorry, I couldn't understand that. Please try again."

    The transcription failed. Speak again, a bit closer to the microphone and with less background noise.

    Nothing happened after I spoke.

    If no speech was detected, the recording is discarded silently. Tap the microphone and try again.

    I see "Voice reply isn't available right now."

    The spoken narration could not be generated for that reply. The full answer is still on screen in text.

    Related guides

    Still stuck? Talk to our team

    Open a support ticket inside the app and our team will answer there.

    Log in and open a support ticket

    Try it yourself, free

    The software is 100% free with unlimited orders. No credit card, no demo, no contract.

    Create your free account