Storytel Revolutionizes Audiobook Narration with ElevenLabs-Based Voice Switcher

According to Storytel's user data, 9 out of 10 customers stop listening to an audiobook not because of the book's content but because the narrator's voice doesn't suit them. However, traditional production methods made it impossible to hire two narrators for a single book due to cost issues.

To solve this problem, Storytel partnered with ElevenLabs to develop the Voice Switcher feature. This feature allows listeners to directly choose between the existing human narrator and an AI-narrated version of the same content.

Particularly targeting the Nordic market, Storytel built a customized model for Nordic languages using its vast proprietary audio data. Starting with Sweden in early 2024, the service has been sequentially expanded to Denmark and Finland, providing high-quality, long-form AI narration.

Currently, Storytel generates narrations through the ElevenAPI, and to improve quality, it goes through internal pre- and post-processing steps such as pronunciation dictionaries and proper noun checks. Going forward, the goal is to minimize manual editing and build a fully automated API-based mass production system.