Voice and AI agents
Voices and recordings
Updated October 1, 2026
Recording Studio in Drop Cowboy® holds your recorded messages and your Mimic AI® voices. You can use either one in a campaign, an automation or an AI agent.
Record a message
Do this when you want a saved audio file to use in a ringless voicemail, a voice broadcast or an automation.
- Open Recording Studio in the left menu and stay on the Recordings tab. Click Add. If you don't see Recording Studio or Add, ask your account owner for access.
- In Add Recording, type a Recording Name and click Continue.
- Pick how to make it:
| Method | What you do |
|---|---|
| Upload a file | Pick an MP3 or WAV file, then click Save Recording. |
| Record now | Click Start Recording, speak, then Stop Recording. Play it under Preview. Click Save Recording, or Record Again. |
| Dial in by phone | Call the number shown, enter the PIN, and record. The panel shows Waiting for your recording... until it arrives. |
| Generate with AI voice | Pick a voice under Mimic AI® Voices or Pro Voices, type up to 500 characters, click Preview, then Save Recording. |
- The recording appears in the Recordings list. To use it, pick it on the message step of a campaign. See Send a ringless voicemail.
If it doesn't work:
- "Microphone access was denied." Allow the microphone in your browser settings and try again, or pick Upload a file.
- "Your browser does not support browser recording." Use Upload a file or Dial in by phone instead.
- "You've edited the script -- preview it again before saving." Click Preview once more, then save.
- The AI voice list is empty. Clone or design a voice first. See Clone your voice.
Clone your voice
Do this when you want campaigns and agents to speak in your own voice from typed text. Only clone a voice that is yours, or one whose owner gave you explicit permission.
- Open Recording Studio in the left menu and click the Mimic AI® Voices tab. Click Add.
- In Add Voice, type a Voice Name and click Continue.
- Click Clone your voice.
- Pick how to give the sample: Record with microphone, Upload a sample, or Call in to record.
- Set Biological Sex to Male or Female. It helps cloning accuracy.
- Check the consent box. It reads "I confirm this is my voice or I have explicit consent from the voice owner to clone it."
- Click Continue, or Get Dial-In Info for the phone option.
- Give 10-30 seconds of clear speech. Samples longer than 30 seconds are rejected.
- Microphone: click Start Recording, read aloud, then Stop Recording. Click Clone Voice, or Record Again.
- Upload: pick an MP3 or WAV file, then click Clone Voice.
- Phone: follow Dial In to Record Your Sample. The panel shows Waiting for your recording...
- The list shows Cloning... for a few minutes. When you see "Voice is ready for text-to-speech generation.", click Try it now.
- Under Text-to-Speech, type a line and click Preview. To keep it as a message, type a Recording name and click Save to Recordings.
Speak in a quiet room at your normal pace. Background noise and music lower the quality of the clone.
If it doesn't work:
- The voice shows Draft. Click Add to Cart and complete checkout. The voice can't be used until you do.
- The voice shows Pending payment. Click Manage Plan to finish checkout. See Choose a plan.
- Continue stays grey. Check the consent box first.
- The sample was rejected. Record or upload a new one between 10 and 30 seconds long.
- The clone doesn't sound like you. Open the voice and click Re-record Sample with a cleaner recording. You see "Re-cloning with your new sample..." while it works.
Choose a voice
Do this when you send a typed message or set up an agent and need to pick who speaks. There are three kinds of voice.
| Voice | Where it comes from | Where you see it |
|---|---|---|
| Pro voice | Ready-made voices you can use right away | Pro Voices, or Pro voices in agent setup |
| Your cloned voice | Clone your voice | My Voices, or Custom (clone) in agent setup |
| Designed voice | A new voice built from a written description | Designed Personas, or Designed in agent setup |
Design a voice with AI
- Open Recording Studio in the left menu and click Mimic AI® Voices. Click Add, type a Voice Name, and click Continue.
- Click Design with AI.
- Write the voice in Voice description, up to 500 characters. Or start from US dialect presets, or open Structured persona and fill in the fields.
- Pick a Language. Add Sample text if you want to hear a specific line.
- Click Generate preview. Click Generate another until you like it, then click Save this voice.
- When the banner "Your AI voice is being designed" clears, the voice is ready to pick.
Pick the voice for a send
- Open Campaigns in the left menu. The page is titled Broadcasts. Start a ringless voicemail or voice broadcast.
- On the message step, click Type it (Mimic AI® voice). To play a saved file instead, click Use a recording.
- Open Select a voice... and pick from My Voices, Pro Voices or Designed Personas.
- Type the message. Check What they'll hear before you continue.
In an automation, add Leave a Voicemail, set Message Type to Text-to-Speech (AI Voice), and pick the AI Voice. See Build an automation.
An AI agent picks its voice on the Voice step of setup. See Set up an AI receptionist.
If it doesn't work:
- A voice is missing from the list. It may still be cloning, or waiting on checkout. Open it in Recording Studio and check its status.
- The list says "Upload or dial in to add a voice". Your team has no cloned voice yet. Clone one, or pick a Pro voice.
This information is for educational purposes only and does not constitute legal advice. Regulations vary by jurisdiction and use case. Always consult with qualified legal counsel to ensure your specific practices comply with applicable federal and state laws.