Who will stop ElevenLabs? After having completely transformed speech synthesis using AI, the company is now launching into the generation of sound effects for cinema and video games.
The start-up is already ahead of all its competitors in terms of speech synthesis. It is to sound what OpenAI (which has just upgraded its chatbot to an omni version) or Anthropic (Claude 3) is to text and already offers powerful AI models to imitate voices. Today, it is tackling a whole new area with the help of a powerful new AI tool capable of generating sound effects at will, just from text.
A new frontier for ElevenLabs
This new tool is soberly called “Text to Sound Effects” and it allows you to create a vast range of effects, short instrumental tracks, soundscapes and character voices from simple text inputs. Designed for audiovisual content creators, whether they work in the film industry, video games or on social networks, this tool aims to facilitate production, therefore saving time and money for those concerned.
To make this progress possible, ElevenLabs has partnered with Shutterstock, a platform well known for its images, but which also offers audio content. It is thanks to the latter’s immense repertoire that ElevenLabs was able to refine its AI model.
Aimee Egan, CEO of Shutterstock said: “ We are excited to collaborate with ElevenLabs to power a major new innovation in AI, Text to Sound Effects, using our ethically sourced data. Combining our rich and immersive library of audio tracks with this cutting-edge technology has created a true market first. We’re excited by the positive feedback from the community having had early access and we can’t wait to see the diversity of projects they will realize “.
A flexible and affordable tool
With user-provided guidance, their new tool can generate up to 22 seconds of audio and offers a minimum of four downloadable audio clip options. Although the tool is free to use, users opting for a paid subscription benefit from commercial licenses for the generated clips, while free users must attribute authorship to ElevenLabs by including “elevenlabs.io” in the title of their production.
Users have a quota of characters allocated for writing indications. Users using it for free benefit from 10,000 characters per month. Regarding sound effects, each second of audio generated consumes 40 characters of the allocated quota, or 200 characters for the default duration of each audio clip request.
Let’s say you want to create 10 second soundtracks without paying the subscription, you can generate 25 per month. To try it out, you can go to the official ElevenLabs website dedicated to Text to Sound Effects.
Download
- Realism of voices
- Management of intonation and emotions
- Many languages supported
ElevenLabs is a benchmark in the sector of voice generation using artificial intelligence. Based on a freemium model, it offers interesting features, particularly those relating to video dubbing in several languages. ElevenLabs has established itself as an essential platform in the field.
ElevenLabs is a benchmark in the sector of voice generation using artificial intelligence. Based on a freemium model, it offers interesting features, particularly those relating to video dubbing in several languages. ElevenLabs has established itself as an essential platform in the field.
Sources: The Verge, Eleven Labs
1