ElevenLabs Voice Generator
ElevenLabs Voice Generator – a script that lets you generate voices from text prompts using the ElevenLabs service. The script acts as a wrapper for the ElevenLabs API, sending requests and importing the generated audio directly into REAPER.
All content generation limits are defined on the ElevenLabs side.
- Install Core modules
- Open Extensions -> ReaPack -> Browse Packages in the menu bar
- Find and install ek_Generate voice via ElevenLabs
- Execute script ek_Generate voice via ElevenLabs in the action list
- Enter licence or start trial for 14 days and restart the script. You can get license key on Gumroad.
- To use the script, you’ll need an API key from Eleven Labs. You can obtain one at: https://elevenlabs.io/app/settings/api-keys
- Create key and enable following access: Text to Speech, Speech to Speech, Voice: Write, Models: Read, User: Read or use no restricted key.
The script has three main modes:
- Simple Text – Generate voice directly from a text prompt.
- Take Text from Item Notes – Use text from item notes as prompts, allowing multi-mode generation. Voices are written into a new take of each item.
- Take Audio Voice from Items – Recognize text from audio waveforms and replace it with a new take using the selected voice.
The service offers a vast library of uploaded professional voices, so you're sure to find something that fits your needs. The script also includes an amazing feature for finding a voice by audio sample. Simply select an item with voice and click the Find Similar button next to the search bar.
ElevenLabs offers several models for voice generation. Each model has its own generation settings — feel free to tweak them to get results that better suit your needs.
Starting with version 3, a more advanced prompt mode was introduced, allowing you to add service phrases and tags to convey emotional nuance. SSML support was also added.
The service also lets you create your own voices, so the script includes a dedicated Voice Manager with two creation modes:
- Design a voice from a text prompt (available with the free plan)
- Clone a voice from an audio recording
Here, you can also manage your voices — edit or delete them.
The script includes a settings popup where you can configure the following options:
- API Key – your key for accessing the service API.
- Output Format – the format in which audio will be generated. Keep in mind that quality depends on your subscription plan. By default, the free plan only supports MP3 at 44.1 kHz, 128 kbps.
- Similarity Threshold – sensitivity for searching similar voices by audio sample. The lower the value, the closer the match.
- Quota Warning – an extra safeguard against overly large requests. If your generation request exceeds the specified number of characters, the script will ask you to confirm before sending it to the API. Zero means no warning
- Remove Background Noise – cleans up background noise before uploading a voice sample for generation. Useful for noisy or low-quality recordings.
- Match Item Length to Generated Content – available in Take Text from Item Notes and Take Audio Voice from Items modes. When enabled, the script adjusts item length to fit the generated audio. If disabled, item length remains unchanged.
Additionally, the script allows you to check your subscription balance directly from REAPER. Simply click Check Balance.
