ElevenLabs, which makes voice-cloning tools, has now announced a new tool by which users can create sound effects based on text prompts. The company had first announced the project in February.

The tool, which will be available for all its users from today, lets one type in prompts like "waves crashing," "metal clanging," "birds chirping," and "racing car engine" to generate snippets of sounds.

The tool of sound effects also allows for creating instrumental clips of up to 22 seconds long from guitar loops, jazz saxophone solos, and music techno loops.

Free users get 10,000 character generations per month-a sound byte generation takes around 150 characters per request. So free-tier users can essentially generate nearly 60 sound effects per month. Also, they have to attribute the sound to "elevenlabs.io" in the title while publishing any content containing the sound clip.

ElevenLabs claimed that it used Shutterstock's audio library, which contains the licensed tracks, as a training tool for its model. The company further indicated that video game developers, film producers, social media content creators, and marketers had also tested the tool during its alpha testing phase.

The startup pointed out that the tool does not produce audio as it accepts prompts that incorporate content and activities which the startup particularly forbids under its Prohibited Content and Uses Policy, including such themes as self-harm and threats to child safety or fraud.

Although there are only a few companies and startups working on AI-based sound generation, the music generation space appears to be crowded for ElevenLabs. There's Dance Diffusion from Stability AI-backed Harmonai; MusicLM from Google; Jukebox from OpenAI; and AudioCraft from Meta. Not to forget, TikTok and Adobe, which have both experimented with generative AI-based music creation tools of their own.