Company
ElevenLabs is an AI voice technology company known for text-to-speech, voice cloning, and speech-to-text tools used in media and conversational AI.
Fish Audio released a 35B mixture-of-experts speech model on Oct 13 that scores 93.2% on Artificial Analysis' pronunciation benchmark. Free commercial use stops at $500,000 in annual revenue, and the two-RTX-3090 setup is one Reddit user's report.
Reality
Andreessen Horowitz partner Olivia Moore published a ranking of the top 100 consumer AI apps that puts ChatGPT far ahead of every rival. In a TechCrunch interview, she argued that growth depends on ads and cheaper models paying for the users who do not subscribe.
Perspective Coverage
3 publishersReality
Prime Minister Andy Burnham is expected to announce a clampdown on non-compete clauses and long notice periods used to hold tech and finance staff. Employers that rely on these terms get a direction to plan against from the speech, but the rules themselves have not been written.
Perspective Coverage
3 publishersReality
Endeavor Catalyst closed its fifth fund at $320 million, taking assets under management past $850 million. For founders outside the big tech hubs it is real money behind two gates: admission to Endeavor's network and a lead investor willing to set the terms.
Perspective Coverage
3 publishersReality
ElevenLabs' $300 million employee tender, led by Wellington and T. Rowe Price, values the London voice AI firm at $22 billion, double its February price. The company raises no new cash, so the doubled price comes from roughly 1.4% of its stock changing hands.
Perspective Coverage
6 publishersReality
Tokyo judge Aya Takahashi ruled that using a performer's voice without permission can infringe publicity rights, in Japan's first AI voice-clone case. Teams shipping synthetic narration there now have to judge whether a voice is being sold on who it sounds like.
Perspective Coverage
4 publishersReality
David AI's 21-language benchmark found speech systems' error rates spread an average of 25.4 points in five Indic languages, against 3.7 in the tightest five. In those languages, choosing a system is a 25-point decision that an English score can make look like a close call.
Reality
ElevenLabs is valued at $22 billion, double its February mark, after employees sold $300 million of shares to investors led by Wellington and T. Rowe Price. Its usage evidence is a self-reported conversation count, and teams sizing voice automation still need completion rates the announcement did not include.
Reality
ElevenLabs released Eleven v4 Turbo for voice agents, reporting medians of about 100 ms inference latency and 150 ms to first speech. Moving an existing agent takes more than a model ID swap, since older voice clones need retraining and SSML break tags no longer work.
Perspective Coverage
3 publishersReality
Synthesia, valued at $4 billion, has an interactive avatar of its corporate affairs head answering common press questions about the company. Comms leads thinking of doing the same need a rule for what a synthetic stand-in may say to outsiders.
Reality
Generate Music, Speech, and Sound Effects are now generally available, and Adobe is selling them on commercial safety. The supplied terms stop short of spelling out who eats a claim.
Perspective Coverage
3 publishersReality
ElevenLabs CEO Mati Staniszewski, whose agents handle Klarna's first-line support for 35 million U.S. customers, says callers should be told when AI answers. He limits the norm to a period when callers still expect a person, and his suggested version makes disclosure a routing choice at the top of the call.
Reality
Universal Music and ElevenLabs announced a multi-year deal on September 10 for a remix and vocal platform still in development, with no launch date, no price and no participating artist named so far.
Perspective Coverage
3 publishersReality
SpaceXAI's new transcription model costs the same as the one it replaces and drops into existing code, so the migration work is re-validating accuracy claims that rest mostly on the company's own test sets.
Reality
Loquent's 14,000 vox-bench runs found Claude Haiku's latency tail at 2.85x its median while ElevenLabs tightened to 1.92x. Alert thresholds tuned to June's data would now be watching the wrong stage of the pipeline.
Reality
ElevenLabs CEO Mati Staniszewski will let gross margins shrink to win share, at a $600 million annual revenue pace and a reported $22 billion valuation. Decagon trained on its voices and now competes with it, so teams buying voice AI should plan to swap the voice itself.
Reality
Breeze TTS 2 scores Elo 1215 and lists at $34 per 1M characters, about a third of Eleven v3's rate. Only the inference code is Apache 2.0, and commercial use of the weights needs a written agreement from BreezeBlue.
Reality
OpenBMB's 2B-parameter model is Apache-2.0, speaks 30 languages, and serves through vLLM's OpenAI-compatible /v1/audio/speech. The real-time factor quoted for it was measured on a different backend than the one the project recommends for production.
Reality
Version 2.0 arrives at the same price as version 1.0. The accuracy gain SpaceXAI reports is concentrated on short multilingual commands. The evaluation sets behind the claim are drawn from production support calls.
Reality
The Sysdig Threat Research Team says stolen cloud keys were checked against ten hosted model services and fronted to buyers, with the compromised tenant paying for every prompt at invocation prices.
Reality
Build · September 16, 2026 · One report1 publisher
Product · September 11, 2026 · Confirmed5 publishers
Product · August 17, 2026 · Confirmed3 publishers
Build · August 17, 2026 · One report1 publisher