What it is
ElevenLabs describes its business as spanning three areas: ElevenAgents for customer experience, ElevenCreative for content creation, and an AI voice generator.[1] Its product pages list text to speech, music, speech to text, voice cloning, dubbing and narration.[1] Those names cover quite different jobs. Text to speech turns written text into spoken audio; speech to text does the reverse, turning recordings into writing. Voice cloning builds a synthetic voice that resembles a particular speaker from samples of that speaker. Dubbing replaces the speech in an existing recording, usually so that it can be heard in another language, and narration applies a reading voice to long text such as an article or a book.
Product lines
ElevenAgents is aimed at customer experience, and the company also points to sales, hiring and marketing operations as places where its technology is used.[4] The common thread is a voice agent: a program that takes spoken input, decides what to do with it, and answers in speech, so that a caller can simply talk instead of working through a keypad menu. ElevenCreative covers the content side, where the same underlying speech technology is used to produce audio for videos, games, advertising and other material rather than to hold a conversation.
The speech models
Eleven v4 launched on 28 September 2026 and is described by the company as its most emotive text-to-speech model; it was released alongside Eleven v4 Turbo, a version built for low latency.[2] "Emotive" here is about how much of a performance — emphasis, pace, pauses — the model carries into the output instead of flattening everything into a neutral reading voice. Low latency is the delay between sending text and getting audio back, which matters far more in a live conversation than in a pre-recorded narration.
ElevenLabs says Eleven v4 was ranked first by Artificial Analysis and was preferred by about 75% of listeners in blind head-to-head tests against competing models.[2] In a blind test, listeners hear samples from two systems without being told which system made which, and choose the better one; that removes some of the effect of brand names, though it still measures preference rather than any objective notion of correctness. These are the company's own characterisations of the results.
Image and video
Image & Video entered beta on 17 November 2025 as a single creative workflow inside the platform.[3] With it, the product aggregates video models including Veo, Sora, Kling, Wan and Seedance, and image models including Nanobanana, Flux Kontext, GPT Image and Seedream.[3] Rather than offering only one model per task, the workflow collects models from a range of developers in one place, so a user can describe what they want, try the available options and keep iterating on pictures and clips without moving between separate services.
Customers and scale
The company's site names Twilio, The Walt Disney Studios, KPN, TELUS Digital, Cisco, Epic Games, NVIDIA, Revolut, Meta, Zoom and Ramp among the companies it serves.[1] The list runs from communications platforms and telecom operators to game studios, banks and film studios, which fits a product family that is sold both as a technical building block and as a finished creative tool.
ElevenLabs ended 2025 with $350 million in annual recurring revenue and passed $500 million in the first four months of 2026.[4] Annual recurring revenue is subscription income scaled up to a full year, a measure that software companies use because it shows the recurring base rather than one-off sales. The company attributes the growth to enterprises deploying voice agents in customer support, sales, hiring and marketing operations.[4] The third close of its Series D brought in BlackRock, Wellington, NVIDIA's NVentures, Santander, Jamie Foxx and Eva Longoria.[4] A Series D is a late-stage funding round, and a third close means the round was reopened after earlier closings to bring in additional investors — here a mix of asset managers, a corporate venture arm, a bank and two individual investors from entertainment.
20262
-
ReleaseElevenLabs launches Eleven v4, its most expressive text-to-speech model
On 28 September 2026 ElevenLabs released Eleven v4, a text-to-speech model it calls its most emotive yet, alongside a low-latency Turbo variant; the company says it topped Artificial Analysis and was preferred by about 75% of listeners in blind tests.
-
ReleaseElevenLabs passes $500M ARR and adds BlackRock, NVIDIA and others as investors
On 5 May 2026 ElevenLabs said it had crossed $500 million in annual recurring revenue and closed a third tranche of its Series D, with BlackRock, NVIDIA's NVentures, Santander and others investing.
Notes and sources
- ElevenLabs — AI Voice Generator & Voice Agents Platformhttps://elevenlabs.io/
- Introducing Eleven v4, our most emotive modelhttps://elevenlabs.io/blog/eleven-v4
- Introducing ElevenLabs Image & Videohttps://elevenlabs.io/blog/introducing-elevenlabs-image-and-video
- ElevenLabs crosses $500M ARR and welcomes new investorshttps://elevenlabs.io/blog/500m-arr-and-new-investors
Cite this page: ElevenLabs. TellWorld, https://www.tellworld.net/products/elevenlabs. Updated 2026-11-07