ElevenLabs Introduces Text-to-Sound Tool

ElevenLabs, a voice cloning startup, introduced its latest tool today, a tool enabling users to create custom sound effects simply by inputting prompts. This project, initially revealed in February, has finally come to fruition.

Accessible to all users from today onwards, this tool empowers individuals to type prompts such as “waves crashing,” “metal clanging,” “birds chirping,” and “racing car engine,” resulting in instantly generated sound snippets.

Moreover, the sound effects tool can produce instrumental musical segments lasting up to 22 seconds. Users can input prompts ranging from guitar loops and jazz saxophone solos to techno music loops, expanding creative possibilities.

Free users are allotted 10,000 character generations per month, with each sound byte generation requiring approximately 150 characters. This allows free-tier users to create nearly 60 sound effects each month. Any published content featuring these sound clips must also attribute the sound to “elevenlabs.io” in the title.

ElevenLabs said it trained its model using Shutterstock’s audio library, which contains licensed tracks. During the alpha testing phase, the tool was utilized by video game developers, film producers, social media content creators, and marketers.

The startup emphasized that its tool strictly follows its Prohibited Content and Uses Policy, prohibiting sound generation prompts related to self-harm, threats to child safety, and fraud.

The competition in AI-powered sound generation is growing. Apart from ElevenLabs, other notable players include Stability AI-backed Harmonai, which recently launched Dance Diffusion. Google has developed MusicLM, while OpenAI offers Jukebox, and Meta presents its AudioCraft model. TikTok and Adobe have also delved into generative AI-based music creation tools, showcasing a diverse innovation landscape in this field.