ElevenLabs Secures $180M Series C Funding - AI Audio Startup

ElevenLabs Secures $180 Million in Series C Funding

ElevenLabs, a prominent startup specializing in AI audio technology, announced Thursday the completion of a Series C funding round, raising $180 million. This investment establishes the company’s post-money valuation at $3.3 billion.

The investment round was co-led by a16z and ICONIQ Growth. Initial reports regarding the funding appeared in TechCrunch, with the final figures largely confirming previous details.

Funding Allocation and Future Development

The newly acquired capital will be strategically allocated towards the continued development of ElevenLabs’ audio tools and expansion of its business operations.

According to Mati Staniszewski, CEO and co-founder, the company’s research is currently focused on enhancing the expressiveness and controllability of its audio AI models.

Focus on Omni-Models and Multimodal Interactions

ElevenLabs is also prioritizing the creation of “omni-models,” which aim to integrate text-based models with its existing audio capabilities. This integration will facilitate more versatile and nuanced multimodal interactions.

Investor interest in ElevenLabs has surged in recent months, driven by two key factors. The first is the widespread excitement surrounding generative AI. The second is ElevenLabs’ emergence as a leading provider of synthetic voice technology.

Widespread Adoption of ElevenLabs Technology

Numerous major publishers, content creators in sectors like media and gaming, and various tech startups are currently utilizing ElevenLabs’ technology. They are leveraging it to enhance their voice and audio features.

This broad adoption has resulted in a highly competitive funding round, attracting participation from a diverse range of prominent investors.

New and Existing Investors

The Series C round includes investments from new participants such as NEA, World Innovation Lab (WiL), Valor, Endeavor Catalyst Fund, and Lunate, an investment firm based in Abu Dhabi.

Existing investors, including Sequoia Capital, Salesforce Ventures, Smash Capital, SV Angel, NFDG, and BroadLight Capital, also contributed to this funding round.

Strategic Backers and Board Changes

ElevenLabs is also gaining support from strategic investors – companies that actively use its technology and are now investing in its future. These include Deutsche Telekom, LG Technology Ventures, HubSpot Ventures, NTT DOCOMO Ventures, and RingCentral Ventures.

Seth Pierrepont, a partner at ICONIQ, will be joining the company’s board of directors. He will work alongside Jennifer Li from a16z and the company’s co-founders.

ICONIQ’s Investment Strategy

ICONIQ has been actively increasing its involvement in the generative AI startup landscape. Last November, the firm co-led a $200 million investment round in Writer, a company focused on written output.

Pierrepont emphasized the importance of audio as a key modality and expressed confidence in ElevenLabs’ potential to become a significant player in the industry.

He highlighted the company’s technological advancements, rapid growth, and the expertise of its founding team.

Future Focus and Recent Growth

As a board member, Pierrepont anticipates focusing on identifying new applications for audio technology and pinpointing optimal target markets.

ElevenLabs’ ability to secure this funding is particularly noteworthy given the current challenges startups face in closing growth rounds. The company’s Series B round, which valued it at $1 billion, was completed just a year ago.

To date, ElevenLabs has raised a total of $281 million in funding.

The Future Development Roadmap

Beyond continued enhancements to its artificial intelligence models, the company intends to allocate the newly acquired funding towards expanding its conversational AI builder. This expansion aims to broaden accessibility to both individual consumers and through strategic partnerships.

The introduction of an AI conversational agent platform last year included the development of a speech-to-text capability. Staniszewski highlighted a strong desire for significant improvements in this specific area.

“Our objective is to achieve a more nuanced understanding of spoken language within conversations. We are actively developing methods to move beyond solely content generation and incorporate robust speech understanding and transcription,” Staniszewski explained. “While some consider speech-to-text a resolved issue, its performance remains inadequate for numerous languages.”

The company believes it can engineer superior speech detection models due to its internal data annotation teams and their capacity for rapid feedback. Further investment will be directed towards this goal.

A key focus will also be on strengthening the creation of AI-driven conversational agents. This includes support for traditional communication methods like telephony and improved integration of diverse knowledge bases. The current funding round reflects this strategy through partnerships with telecommunications companies.

Customers are already leveraging the platform to access and interact with their existing data archives. For instance, ElevenLabs collaborated with TIME magazine last year to implement a conversational bot enabling users to pose questions about the TIME Person of the Year.

Staniszewski envisions a wider deployment of conversational AI agents across various websites. On news platforms, for example, users could query articles or request summaries from the bot.

The CEO also acknowledged that, despite advancements in AI-powered voice bots, achieving truly natural-sounding interactions that respond appropriately to human emotion and conversational cues remains a challenge.

“The manner in which I communicate with you influences your reaction. Variations in tone – excitement, calmness, or even interruption – should elicit corresponding responses. Current AI solutions are approaching human-level performance, but still fall short,” Staniszewski stated.

Pierrepont of ICONIQ emphasized that effective machine communication hinges on accurate understanding of human speech. If AI fails to comprehend the user, engagement quickly diminishes.

ElevenLabs has primarily expanded its reach and revenue through B2B collaborations. However, the company is also pursuing direct consumer engagement.

In 2024, the startup launched ElevenLabs Reader, its first direct-to-consumer product, an application designed to read articles, text, and documents aloud. Subsequently, the company added functionality to generate podcasts using AI voices from documents and web pages, similar to Google’s NotebookLM. Expanding into further consumer-facing applications is a priority.

Evidence suggests this expansion is already underway. TechCrunch reported that ElevenLabs Reader is currently testing a program allowing users to publish audiobooks directly on the platform. Future plans include providing creators with tools to utilize multiple voices for audiobook narration and enhancing localization capabilities.

Staniszewski indicated the company is exploring methods to improve content distribution for both users and businesses, potentially including its own application. The implications of this move, particularly regarding potential competition with existing customers, are being carefully considered. ElevenLabs currently provides voice technology for platforms such as Pocket FM and Kuku FM.

Currently, ElevenLabs powers AI-generated audio for a range of products and platforms, including Perplexity, Rabbit R1, Chess.com, ESPN, Lex Fridman podcast, The Atlantic, and Synthesia. The company’s overarching goal is to increase its presence across various platforms and establish a comprehensive conversation stack to deliver enhanced experiences and insights to its clientele.

Safety Considerations

While ElevenLabs offers significant advancements, concerns regarding misuse have emerged. The technology has been linked to several instances of misinformation dissemination. A recent analysis by Recorded Future, a threat intelligence firm, revealed the platform's involvement in a Russian propaganda effort.

Previously, the company’s voice cloning capabilities were exploited to generate a convincing audio deepfake of President Joe Biden. In 2023, reports surfaced indicating that users on 4chan utilized the AI audio generation tool to replicate the voices of prominent figures like Joe Rogan, Ben Shapiro, and Emma Watson, subsequently employing these clones to distribute questionable material.

ElevenLabs has demonstrated a proactive approach to addressing these issues. Currently, a policy is in place that explicitly forbids “unauthorized, harmful, or deceptive impersonation.” Furthermore, a combination of automated systems and human review processes are employed to identify and remove inappropriate content.

However, the expansion of the company’s toolset and increased direct engagement with consumers inevitably creates more avenues for malicious actors to attempt exploitation. This necessitates ongoing vigilance and adaptation.

“Being a leader in AI audio technology carries a responsibility to develop robust safety measures alongside the technology itself,” stated Staniszewski. “We consistently prioritize safety, even if it means delaying deployment or foregoing immediate commercial gains.”

Staniszewski further explained that ElevenLabs adheres to the C2PA standard, which utilizes metadata to track content origin. Additionally, the company provides a publicly accessible tool enabling verification of whether an audio file was created using ElevenLabs technology, leveraging digital signatures embedded during the generation process.

This verification method is expected to evolve in response to increasingly sophisticated techniques employed for misuse. Continuous development and refinement of safety protocols are crucial.

Stay informed with TechCrunch’s AI newsletter! Subscribe here to receive weekly updates every Wednesday.