ElevenLabs, an AI audio research and development startup, said on Tuesday it supplied AI voice technology to global audio platform company SpoonLabs, cutting the production time for audio novel content from several months to a few hours.

SpoonLabs redesigned its production approach as it expanded from a live-audio-focused business into story-based audio content. Under the traditional voice-actor recording method, producing a single piece of content took 4 to 7 months.

Before adoption, SpoonLabs tested multiple text-to-speech (TTS) solutions at home and abroad under conditions similar to its actual production environment. It evaluated changes in intonation based on punctuation and the ability to express emotions based on context as key criteria, and found ElevenLabs scored highest among the comparison group. Another factor was that it provides functions needed for audio production, including voice cloning and background music and sound effects generation, on a single platform.

After adopting ElevenLabs technology, SpoonLabs launched its audio novel service PodNovel simultaneously in 3 countries in January, with 30 titles in South Korea, 26 in Japan and 19 in Taiwan. SpoonLabs plans from this month to release at least 3 new pieces of content per country each week to build a lineup of more than 100 titles in the short term.

Kim Hyun (김현), head of the PodNovel content team at SpoonLabs, said, "ElevenLabs provided acting-level technology that understands context and emotions." He said, "AI-based production has dramatically improved production speed and scalability, and this is a shift in the production method itself."

Hong Sang-won (홍상원), head of the ElevenLabs Korea branch, said, "Collaboration with SpoonLabs allowed us to fundamentally improve the way audio content is produced." He said, "We will work with various media companies in the future to create a new production standard."