Skip to content

Elevenlabs, the Hot Audio Startup, confirms $ 180 million in Funds of the C series with an assessment of $ 3.3b

ElevenOne of the new most popular companies that work in the AI ​​Field audio, said Thursday that he has raised a round C of $ 180 million, valuing the company at $ 3.3 billion after money. The growth of A16z and ICONIQ are a co-leader investment.

The rumors of fundraising were First reported by Techcrunch. The final numbers confirm some, but not all the details that we report previously (specifically, the general size of the round is smaller than we had heard; the valuation and the main investors are the same).

The financing will be used to continue building the Elevenlabs audio tools and for business development.

Mati Staniszewski, the CEO that co -founded the company with the childhood friend Piotr Dabkowski, said in an interview that the startup is focusing its research on the creation of audio AI models that are more expressive and have more control. Staniszewski added that the company is also focusing on the “OMNI models”: combining text -based models with its audio models for multimodal interactions.

There has been a frenzy of interest in investors in eleven Returning several monthsthoroughly of two important currents. First, there has been a great wave of exaggeration around the generative AI that has been catching many companies in its path. Secondly, Elevenlabs has become an important player among those who provide synthetic voice technology. Dozens of important editors and creators of vertical content such as media and games, as well as several other new technology companies, are using Elevenlabs technology to feed their voice and audio characteristics.

As expected, that has translated into a very full financing round with many outstanding names.

The new investors in this C series include Nea, World Innovation Lab (Wil), Valor, Endeavour Catalyst Fund and Abu Dhabi Investment Firm Lunate. The previous investors also participate include Capital Sequoia, Salesforce Ventures, Smash Capital, SV Angel, NFDG and Broadlight Capital.

Together with these, Elevenlabs is also collecting several new strategic sponsors, that is, companies that use their technology that is now also investing in you as well. These include Deutsche Telekom, LG Technology Ventures, Hubspot Ventures, NTT Docomo Ventures and Ringcentral Ventures.

ICONIQ partner, Seth Pierrepont, will join the company’s board, along with the existing Jennifer Li Board members of A16z and the company’s co -founders.

Iconiq has increased its activities around new generative companies of AI. Taking advantage of the written departure, the company too co-lided a $ 200 million round in writer last November.

“We have always felt that audio is a very important modality, and we think there will be a very large company built in this category,” Pierrepont told TechCrunch. “We have observed eleven OnLabs of its launch, and we were impressed by the quality of technology, how fast it rose in terms of mentality and impulse, and the depth of the domain experience of the founders.”

Pierrepond added that, as a member of the Board, many of the conversations with the company will be around creating new use cases for audio and finding the appropriate markets for it.

At a time when new companies are still difficult for them to close growth rounds, it is remarkable that Elevenlabs raised their series B of $ 80 million, which valued it at $ 1 billion, Just a year ago. Elevenlabs has raised a total of $ 281 million to date.

The product roadmap

In addition to an approach in improving its AI models, the company plans to use funds to increase its conversational builder with the ambition to reach more consumers directly and through associations.

Piotr Dabkowski and Mati Staniszewski co -founders. Image Creidts: Elevenlabs

Last year, the company debuted A AI conversation agent platformAnd a key part of that product was to develop a voice component to text. Staniszewski said the company wants to improve in that area much more.

“We want to understand what you say best in a conversation. We are working on ways to get away from just generating content and understanding and transcribing speech, ”said Staniszewski. “Many people say that speech is a problem solved. But for many languages, it is quite bad. We believe that we can build better voice detection models because we have internal equipment to write down data and give us quick comments. “

The company also wants to double the creation of conversation agents with AI when supporting inherited communications such as telephony and better integrate different types of knowledge sources. This is partly that is being associated with telecommunications in this round.

It is also being used by its customers to take advantage of their own files. Last year, ElevenLabs associated with Time publication to implement a conversational bot For users to ask questions about the time of the year.

Staniszewski said the company imagines more conversational agents in the sites: at news sites, for example, users could ask questions about stories or ask the bot to summarize them.

The CEO also pointed out that while the quality of voice bots with AI has improved, the problem of sounding natural while reacting to humans who speak or emotion in different ways has not yet been resolved.

“The way I talk to you impact how you react or answer me. Sometimes, I will be excited, or sometimes, I will be calm and sometimes I will interrupt you. You will answer me accordingly. The current generation’s solutions are about to be good, but they are not as good as humans, ”said Staniszewski.

ICONIQ Pierrepont also emphasized that if AI does not understand it well when talking, the communication of the machine breaks down and users immediately lose interest.

Elevenlabs has mainly increased its scope (and revenue fun) through B2B associations. But it is also coming out in a direct limb.

In 2024, the startup launched its first purely consumer oriented product, Elevenlabs Reader, An application that reads articles, text and documents. Later, the company added the ability to Create a podcast with generative voices Of documents and web pages, not very different from what you can do with the Google notebook. Staniszewski said he wants to expand to more experiences of consumers.

You can actually be doing that. Techcrunch saw that the company has been testing a program in the Elevenlabs Reader application that invites users to publish audiobooks on the platform. The company also wants to give tools to creators so that multiple voices read an audiobro in the future and at the same time create a better location.

Staniszewski said the company is discovering ways for users and companies to better distribute their content, even in their own application. If that takes him to a real direct competition with his clients it will be something to see. (That has been a reason why many B2B technological companies prefer to stay away from direct reproductions to consumers). In particular, Elevenlabs enhances voice technology for audio content platforms such as Pocket FM pocket with light speed and Kuku FM backed by Google.

Elevenlabs already could audio generated by AI in products and platforms such as perpleity, Rabbit R1, Chess.com, ESPN, Lex Fridman Podcast, The Atlantic and Synthesia. The company’s objective is to be in more places and also have a pile of end -to -end conversation so that it can generate more experiences and ideas for its customers.

Security

Not all ElevenLabs silver coatings have been without clouds: their technology has been involved in some remarkable erroneous information campaigns. A recent report by the Threat Intelligence Company Future found that The company’s product was used in a Russian propaganda operation. Last year, someone used the company’s voice platform to create A deep audio of Joe Biden. In 2023, Motherboard They reported that 4Chan members supposedly used the AI ​​Audio Generation tool to create voices that sounded like Joe Rogan, Ben Shapiro and Emma Watson to disseminate problematic content.

But the company has responded quickly. Today, it has a policy that prohibits “unauthorized, harmful or misleading supplantation.” In addition, it uses a mixture of human moderation directed by machine and to eliminate this content. However, as the company increases its set of tools and has more direct consumption points, this opens the door to more opportunities so that malicious actors look for ways to use it badly.

“As one of the favorites of AI audio work, we treat it as our responsibility to develop the appropriate security mechanism as we build technology. We will often make decisions to prioritize the safety of the speed of implementation or commercial benefit, ”said Staniszewski.

Staniszewski added that although the company follows C2PA, a standard to track content with metadata, it also has A public tool That allows anyone to verify if the audio was generated through Elevenlabs technology using digital firms that place in the audio during the generation. That could also be a clue that continues to develop over time, since misuse approaches also become more sophisticated.

Leave a Reply

Your email address will not be published. Required fields are marked *