ElevenLabs
ElevenLabs is an AI voice generation platform offering text-to-speech, voice cloning, and conversational AI agents in 70+ languages. The platform uses deep learning to produce lifelike synthetic speech for creators, developers, and enterprises, with a celebrity voice licensing marketplace.
Score generated by AI agents based on publicly cited evidence and reviewed by the project maintainer. Not independently validated. Last assessed 2026-10-23. Score revised 2026-10-23: 41 → 30.
Score History
Timeline events are AI-curated from public reporting. Score trajectory is derived from documented events.
ElevenLabs opened its text-to-speech and voice-cloning beta in January 2023 with a generous free tier. Within a week 4chan users had used it for celebrity hate-speech deepfakes, and the company proposed payment or ID checks for cloning. A $19 million Series A in June came with a free AI Speech Classifier, and in August the company exited beta with Multilingual v2 and Professional Voice Cloning. Quality drove growth, pricing was simple, and misuse was the main problem.
The $80 million Series B at a $1.1 billion valuation opened a year of expansion: Voice Actor Payouts, the Reader app, the Omnivore team hire and the Conversational AI platform. Days after the round, the Biden robocall that experts said was likely made with ElevenLabs brought national scrutiny, and in August voice actors sued over default voices they said imitated them. The company answered with account bans, a Reality Defender partnership and AI safety work, while pricing stayed straightforward.
The February 28, 2025 terms and privacy update granted ElevenLabs a perpetual license over voice recordings and derived models. Kukarella ended its partnership over it, and in March Consumer Reports found ElevenLabs' consent checks relied on self-attestation. Through 2025 the product kept improving (Eleven v3, and Eleven Music licensed from Kobalt and Merlin) while the ways to charge users multiplied: ElevenReader hour-based plans, model-specific credit consumption, the Iconic Voice Marketplace and v1 deprecation. Vacker settled in August and was dismissed in November, and a $100 million employee tender valued the company at $6.6 billion.
The $500 million Series D at $11 billion, with Sequoia's Andrew Reed joining the board, began a scale-up toward an IPO. ARR passed $500 million by May, a former Adyen CFO joined in September, and a $300 million employee tender valued the company at $22 billion. Users got price cuts (Text to Speech API up to 55% cheaper, Music up to 50%), Eleven v4 and a Universal Music Group licensing deal. They also got tier reshuffles that moved features off the Free plan and forced migrations from legacy models. A May 2026 Illinois biometric-privacy class action revived the question of how ElevenLabs obtained the voices it trained on.
Alternatives
AI voice generator focused on content creators with 200+ natural-sounding voices and an intuitive studio interface. Moderate switch — re-upload scripts and re-generate audio. Creator pricing is advertised at $19/month when billed annually ($228/year). Smaller voice library but sufficient for most use cases.
AWS text-to-speech service with pay-per-use pricing (no credits or confusing tiers) and deep AWS ecosystem integration. Easy switch for developers already on AWS. Voice quality is good but less natural than ElevenLabs for creative use cases. Best for high-volume, cost-predictable production workloads.
Dimensional Breakdown
Summaries below were written by AI agents based on the cited evidence. They are editorial interpretations, not independent research findings.
Dimension History
Timeline (50 events)
Beta platform launches publicly with voice cloning
ElevenLabs publicly released its beta text-to-speech and voice cloning platform in January 2023. The initial suite included synthetic voices and voice cloning from short audio samples. The platform gained momentum rapidly, praised for voice output quality and a generous free tier.
4chan users weaponize voice cloning for hate speech
Within one week of ElevenLabs' beta launch, 4chan users exploited the platform to generate celebrity deepfakes, including an Emma Watson-sounding voice reading Mein Kampf and a Ben Shapiro-sounding voice making racist remarks about Alexandria Ocasio-Cortez. Joe Rogan-sounding voices also appeared among the abusive clips. ElevenLabs acknowledged 'an increasing number of voice cloning misuse cases' and proposed requiring payment information or ID verification for cloning access.
Series A and AI Speech Classifier released
ElevenLabs raised $19 million in Series A funding co-led by Andreessen Horowitz, Nat Friedman, and Daniel Gross, valuing the company at approximately $100 million ($99 million post-money, according to TechCrunch). The company also announced its free AI Speech Classifier for detecting ElevenLabs-generated audio and reported more than one million registered users. Contemporary coverage reported a claimed 99% accuracy on unmodified audio, falling to about 90% after some modifications.
Exits beta with Multilingual v2 and Professional Voice Cloning
ElevenLabs officially exited beta and released Eleven Multilingual v2, a foundational model supporting nearly 30 languages while maintaining a speaker's vocal characteristics across languages. The release followed the public availability of Professional Voice Cloning to all creators, which ElevenLabs said could produce a virtually indistinguishable digital copy of a user's own voice. Contemporary reporting described a minimum of 30 minutes of audio, with three hours optimal.
AI Dubbing tool launches with 20+ language support
ElevenLabs released AI Dubbing, a tool translating speech into 20+ languages while preserving the speaker's original voice, emotions, and intonation. The feature combined multilingual speech synthesis, voice cloning, noise removal, and speaker differentiation into a single pipeline, targeting independent creators who lacked resources for professional localization.
Series B raises $80M at $1.1B unicorn valuation
ElevenLabs raised $80 million in Series B funding co-led by Andreessen Horowitz, Nat Friedman, and Daniel Gross, with Sequoia Capital and other investors participating. CEO Mati Staniszewski said the round valued the company at $1.1 billion, making it a unicorn less than two years after its 2022 founding. The rapid valuation escalation from approximately $100 million in June 2023 increased the scale of its venture backing.
Biden deepfake robocall targets New Hampshire primary voters
AI-generated robocalls impersonating President Biden urged New Hampshire voters not to vote in the January Democratic primary. Pindrop and University at Buffalo expert Siwei Lyu concluded the recording was likely made with ElevenLabs. According to a person familiar with the matter, ElevenLabs suspended the creator's account; the company itself declined to comment on a specific incident. Its classifier initially returned a 2% match, rising to 84% on audio cleaned by Pindrop. The later charges and FCC fine are recorded separately in the May 2024 event.
Voice Actor Payouts let creators earn from shared voice clones
ElevenLabs added Voice Actor Payouts to its Voice Library, letting users share verified Professional Voice Clones and earn money when paid subscribers use them. Creators could take rewards in US dollars, paid through Stripe Connect, or in ElevenLabs credits.
Selected for 2024 Disney Accelerator program
ElevenLabs was selected for the 2024 Disney Accelerator, a business-development program connecting emerging companies with Disney's expertise. Disney said each participating company receives guidance from its senior leadership team. The selection signaled mainstream entertainment industry interest in ElevenLabs' voice technology.
FCC proposes $6M fine against Steve Kramer over Biden deepfake robocalls
The FCC proposed a $6 million fine against political consultant Steve Kramer, who admitted commissioning the AI-generated Biden robocall sent to thousands of New Hampshire voters. It separately proposed a $2 million fine against Lingo Telecom, accused of transmitting the call. Kramer was indicted in four New Hampshire counties on 13 felony voter-suppression counts and 13 misdemeanor counts of impersonating a candidate.
Reader app launches on iOS, later adding licensed Iconic Voices
ElevenLabs launched its Reader app in June 2024 on iOS, reading articles, PDFs, newsletters, e-books and other text aloud in English. By July 2 it had added licensed voices of Judy Garland, James Dean, Burt Reynolds, and Sir Laurence Olivier through CMG Worldwide. Android availability and a worldwide rollout in 32 languages followed later that summer.
Partners with Reality Defender on deepfake detection
ElevenLabs announced a July 2024 partnership with Reality Defender to support audio-deepfake detection for governments and enterprises. In an April 2, 2025 retrospective, Reality Defender said it had added more than 295 hours of synthetic voice data to its datasets and achieved a tenfold improvement in data-generation efficiency: generating and processing data at least ten times faster than with alternative synthetic voice platforms.
Voice actors file landmark AI cloning lawsuit
Voice actors Karissa Vacker and Mark Boyett, authors Brian Larson and Vaughn Heppner, and publisher Iron Tower Press filed suit against ElevenLabs. The complaint alleged that default voices 'Bella' and 'Adam' imitated Vacker and Boyett respectively without consent or compensation, and that ElevenLabs used copyrighted audiobook material in training. It raised publicity-rights and Digital Millennium Copyright Act claims.
FCC finalizes $6 million fine against Steve Kramer
The FCC adopted a $6 million fine against political consultant Steve Kramer for illegal robocalls using an AI-generated clone of President Biden's voice and caller ID spoofing aimed at New Hampshire voters before the January 2024 primary. The fine targeted Kramer, not ElevenLabs.
Omnivore reader-app team joins ElevenLabs
The team behind Omnivore, an iOS reader app used by 500,000 people globally, joined ElevenLabs to expand ElevenReader and audio accessibility. ElevenLabs said Omnivore's open-source codebase would remain publicly available, while its team moved to work on ElevenReader.
Conversational AI platform launches for voice agents
ElevenLabs released its Conversational AI developer platform, enabling businesses to build interactive voice agents with customizable tone, response length, and knowledge bases. The SDK supported Python, JavaScript, React, and Swift. The product expanded ElevenLabs from a TTS tool into voice-agent infrastructure, adding a service billed by conversation duration.
Credit system restructured with model-specific quotas
Flexprice’s retrospective guide reports that in early 2025 ElevenLabs divided character usage into separate Multilingual v2 and Conversational v1 credit buckets with different overage rates, increasing pricing complexity. Fieldwork dates the restructuring to January 2025. Different services also consumed credits at varying rates; the historical quota account lacks contemporaneous primary corroboration.
Series C raises $180M at $3.3B valuation
ElevenLabs closed a $180 million Series C round co-led by Andreessen Horowitz and ICONIQ Growth, with new investors including NEA, World Innovation Lab, Valor, Endeavor Catalyst Fund, and Lunate. Existing investors including Salesforce Ventures increased their support. The round valued the company at $3.3 billion, tripling its valuation in one year.
February 2025 voice-data terms draw partner privacy concerns
Kukarella reported that ElevenLabs updated its Terms of Service and Privacy Policy on February 28, 2025. It objected to a perpetual, irrevocable, royalty-free, worldwide license over voice recordings and models derived from them, interpreting this as allowing continued use of voice-derived technology after account deletion. Kukarella subsequently announced that it had ended its partnership over these concerns.
Consumer Reports finds voice cloning safeguards inadequate
Consumer Reports assessed six voice-cloning platforms and found that ElevenLabs, Speechify, PlayHT, and Lovo lacked meaningful barriers to non-consensual cloning. They required only a checkbox or similar self-attestation confirming the legal right to clone. Descript required a recorded consent statement; Resemble AI allowed a first clone without verification but required recorded consent for subsequent clones. Consumer Reports warned that such tools could enable scams impersonating relatives.
Kukarella announces it ended its ElevenLabs partnership over ToS changes
Kukarella announced on March 12, 2025 that it had ended its partnership with ElevenLabs the previous week. It cited the February 28 Terms of Service and Privacy Policy updates, saying they created unacceptable risks for users and conflicted with its commitment to protecting digital identity.
Eleven v3 launches with 70+ languages and audio tags
ElevenLabs released Eleven v3 (alpha), a TTS model supporting 70+ languages with expressive audio tags ([whispers], [sighs], [excited]) and multi-speaker dialogue. ElevenLabs positioned it as an expressiveness advance over Multilingual v2. An 80% discount applied through the end of June 2025 for self-serve users generating through the UI. Workflows adopting its model-specific tags and dialogue features could require adaptation when moving to competitors.
ElevenReader introduces subscription plans and hour-based generation
ElevenReader introduced subscription plans and began counting listening hours when it generates audio from users' own content such as articles, PDFs or ePub files. Replays did not use additional hours, one-time hour packs were offered, and a free plan remained.
Steve Kramer acquitted in New Hampshire AI robocall case
A Belknap County Superior Court jury acquitted political operative Steve Kramer on all charges over the robocalls that used an AI-generated voice mimicking President Biden. The verdict concerned Kramer, not ElevenLabs.
Credit system unified across models, according to retrospective pricing guide
Flexprice’s retrospective pricing guide reports that later in 2025 ElevenLabs unified character credits across models, giving each tier a single credit allowance together with defined model access, voice slots, and dubbing entitlements. Different services continued to consume credits at different rates. Fieldwork dates the unification to August 2025; no contemporaneous ElevenLabs announcement was located.
Eleven Music launches with Kobalt and Merlin licensing deals
ElevenLabs launched Eleven Music after striking licensing partnerships with Kobalt and Merlin. Music Business Worldwide reported that the initial tier used production music; a forthcoming Eleven Music Pro model would use cleared Kobalt and Merlin catalog under opt-in deals. Kobalt's agreement provided approximate 50/50 royalty parity between publishing and recorded music revenues, which the outlet described as potentially precedent-setting. Merlin represented more than 30,000 copyright holders through roughly 500 labels and distributors.
Vacker v. ElevenLabs parties report mediated settlement
The Vacker v. ElevenLabs parties told the court on August 20 that mediation had produced a settlement agreement. They agreed to pause proceedings through September 18 to finalize the long-form agreement. Wilftek described this and the Anthropic settlement as the first AI copyright infringement settlements, with 46 other cases then pending. Terms were not publicly disclosed, and no final ruling resolved the underlying voice-rights and DMCA issues.
ElevenLabs launches $100 million employee tender at $6.6 billion valuation
ElevenLabs launched a $100 million tender offer giving employees liquidity at a $6.6 billion valuation, double its Series C valuation from nine months earlier. Existing investors Sequoia and ICONIQ led it.
V1 model deprecation announced for December 2025
ElevenLabs announced that eleven_monolingual_v1 and eleven_multilingual_v1 were deprecated and scheduled for removal on December 15, 2025. It advised users to migrate to newer models for continued service. This was an announced schedule, rather than confirmation of completed removal.
ElevenReader halves Ultra price and adds free monthly listening
ElevenReader cut its Ultra plan to $11 per month or $99 per year, which it described as 50% off the previous price, and gave the Free plan 10 hours of AI audio generation each month.
Vacker v. ElevenLabs dismissed with prejudice after settlement
The District of Delaware granted the parties' joint stipulation of voluntary dismissal with prejudice, closing the voice actors' case after their August settlement. Settlement terms were not disclosed and no ruling decided the underlying claims.
Iconic Voice Marketplace launches with celebrity licensing
ElevenLabs unveiled the Iconic Voice Marketplace, enabling brands to license approved AI voices including Michael Caine, Maya Angelou, Alan Turing, and J. Robert Oppenheimer. Matthew McConaughey invested and used ElevenLabs for a Spanish-language edition of his newsletter; his voice was not available in the marketplace at launch. The marketplace required approval from talent or authorized rights holders for projects including film, TV, advertising, gaming, and audiobook narration.
Series D raises $500M at $11B valuation, eyes IPO
ElevenLabs closed a $500 million Series D led by Sequoia Capital with participation from a16z, ICONIQ, Lightspeed Venture Partners, Evantic Capital, BOND and other investors, valuing it at $11 billion. The company said a16z quadrupled down and ICONIQ tripled down. It closed 2025 with over $330 million ARR; Reuters reported CEO hopes to double that in 2026, implying roughly $660 million. The company was building toward an IPO; its more specific two-to-three-year readiness target was reported in March 2026. Nvidia had invested in September 2025, rather than being named in this round's initial announcement.
ElevenLabs and UK AI Security Institute agree three-year research partnership
ElevenLabs agreed a three-year partnership with the UK Government's AI Security Institute to study the societal and security implications of AI voice, including how well people can tell when they are speaking to an AI agent.
Senator Hassan asks ElevenLabs about voice-cloning scam safeguards
Senator Maggie Hassan, Ranking Member of the Joint Economic Committee, asked ElevenLabs, LOVO, Speechify and VEED what they do to deter scammers from abusing their voice-cloning tools. It was an information request, not an enforcement action.
ElevenLabs reprices plans and reallocates credits
PricingSaaS recorded a restructuring of ElevenLabs' plans between April 13 and April 20, 2026. Starter rose from $5 to $6 a month, Creator credits rose from 100k to 121k and Pro from 500k to 600k, Scale fell from $330 to $299 with credits cut from 2M to 1.8M, and Business fell from $1,320 to $990 with credits cut from 11M to 6M.
ElevenLabs reports ARR above $500 million
ElevenLabs said it ended 2025 with $350 million in annual recurring revenue and passed $500 million ARR in the first four months of 2026. It announced the investment as the third close of its Series D.
ElevenLabs cuts API and Agents prices and adds pay-as-you-go
ElevenLabs cut prices for self-serve ElevenAPI and ElevenAgents customers: Text to Speech up to 55% lower, Speech to Text up to 45% lower and ElevenAgents up to 20% lower. It also introduced pay-as-you-go billing.
Voice professionals file Illinois biometric privacy class action
Seven voice professionals filed a proposed class action against Eleven Labs Inc. in the Northern District of Illinois (Case 1:26-cv-05437). The complaint alleges the company extracted voiceprints from their publicly distributed recordings without the notice and written consent the Illinois Biometric Information Privacy Act requires, and used them to train commercial voice models. These are allegations, not court findings.
Voiceover Studio sunset in favor of ElevenCreative Studio
ElevenLabs' documentation says Voiceover Studio was sunset on May 15, 2026 and directs users to ElevenCreative Studio for new projects.
Voice Library creators pass $22 million in earnings
ElevenLabs reported that voice creators on its Voice Library had earned over $22 million, double the $11 million reported in November 2025, and that more than 10,400 creators were earning on the platform.
Music v2 launches with lower music pricing
ElevenLabs launched Music v2 and cut Music v1 and v2 pricing by up to 50% for ElevenAPI and up to 40% for ElevenCreative self-serve customers.
Legacy v1 speech models rescheduled for July 9 removal
ElevenLabs' changelog said the eleven_monolingual_v1 and eleven_multilingual_v1 text-to-speech models and the scribe_v1 speech-to-text model were deprecated and would be removed on July 9, 2026, replacing the earlier December 2025 schedule. Users were told to migrate to current models before removal.
UK memorandum and plan to double UK headcount
ElevenLabs signed a memorandum of understanding with the UK Department for Science, Innovation and Technology on voice AI for public services, AI security and talent. It said it would double UK headcount to 200 this year and move into a London headquarters three times the size of its existing office.
Image & Video moved off the Free plan
PricingSaaS recorded that Image & Video was removed from ElevenLabs' Free plan, which then showed Image only, and moved to Starter. Between its March 30 and June 29 captures, 192kbps audio also moved from the Creator to the Pro tier.
Court stays discovery in BIPA case pending motion to dismiss
In Amer v. Eleven Labs Inc., the Northern District of Illinois stayed all discovery pending a ruling on ElevenLabs' August 17 motion to dismiss for lack of jurisdiction and failure to state a claim, citing the number, complexity and novelty of the claims.
Former Adyen CFO Ethan Tandowsky joins as CFO
ElevenLabs hired Ethan Tandowsky, most recently CFO of Adyen, as chief financial officer. The company said it had 800 employees and that large organizations made up 55% of its revenue, up from 40% a year earlier.
Universal Music Group and ElevenLabs sign multi-year licensing agreement
Universal Music Group and ElevenLabs announced a multi-year licensing agreement and strategic collaboration to build AI audio products for artists and songwriters and a licensed AI music platform for fans. ElevenLabs called it its first agreement with a major label. No exclusivity or financial terms were stated.
Eleven v4 and v4 Turbo launch
ElevenLabs launched Eleven v4, which it called its most emotive text-to-speech model, and the low-latency Eleven v4 Turbo, with support for Professional Voice Clones, across ElevenAgents, ElevenCreative and ElevenAPI.
$300 million employee tender values ElevenLabs at $22 billion
ElevenLabs closed a $300 million employee tender offer valuing it at $22 billion, double its February Series D valuation. Wellington and T. Rowe Price led the tender.
Evidence (67 citations)
D1: User Value Erosion
D2: Business Customer Exploitation
D3: Shareholder Extraction
D4: Lock-in & Switching Costs
D5: Twiddling & Algorithmic Opacity
D6: Dark Patterns
D7: Advertising & Monetization Pressure
D8: Competitive Conduct
D9: Labor & Governance
D10: Regulatory & Legal Posture
Scoring Log (6 entries)
Checked all 63 timeline/evidence/milestone items plus three alternatives and all prose/identity fields. {'itemsChecked': 66, 'verified': 9, 'corrected': 30, 'resourced': 24, 'removed': 3, 'dateFixes': 10, 'proseFixes': 31, 'inventedDetails': 7, 'unsupportedDetails': 10}. Contradicted specifics: $791M total funding — $781M in the official February 4, 2026 Series D announcement (https://elevenlabs.io/blog/series-d); Seven funding rounds — Five in the same official announcement (https://elevenlabs.io/blog/series-d); $6M fine assigned to Lingo — NPR assigns $6M to Kramer and $2M to Lingo (https://www.npr.org/2024/05/23/nx-s1-4977582/fcc-ai-deepfake-robocall-biden-new-hampshire-political-operative); Tenfold detection efficiency — Reality Defender reports data-generation efficiency (https://www.realitydefender.com/insights/elevenlabs-reality-defender-voice-deepfake-detection); 30,000+ independent labels — MBW reports 30,000 copyright holders through approximately 500 labels/distributors (https://www.musicbusinessworldwide.com/data/merlin-approximate-number-of-annual-members/); Matthew McConaughey listed as a marketplace voice at launch — Variety explicitly says his voice was not part of the marketplace (https://variety.com/2025/digital/news/matthew-mcconaughey-michael-caine-ai-voice-elevenlabs-1236574041/); ElevenLabs article attributed to Security Magazine — The actual article and title concern DeepSeek (https://www.securitymagazine.com/articles/101374-dangers-of-deepseeks-privacy-policy-data-risks-in-the-age-of-ai). Qcall 2.8x, Sacra 5x, 295 training hours, 50/50 music royalty split and $660M ARR target retained with attribution. No score changes.
41→30. Since Feb 2026: $500M Series D at $11B (Feb 4) and $300M employee tender at $22B (Sep 30); ARR >$500M (May); API/Agents price cuts up to 55% and pay-as-you-go (May 7) and Music price cuts (May 26); April 20 plan repricing/credit reallocation and Image & Video moved off Free (Jun 29); v1/Scribe v1 removal rescheduled to Jul 9; Voiceover Studio sunset; Eleven v4 (Sep 28); UMG licensing deal (Sep 10); Illinois BIPA class action (May 11; discovery stayed Aug 26 pending motion to dismiss); Senator Hassan inquiry (Apr 16); UK AISI partnership and DSIT MoU; hiring expansion to ~800 staff, no layoffs; ex-Adyen CFO hired. D1 4→3 (recalibration: product improving, only minor paywall shifts, fits 2-3); D2 4→3 (event: May 2026 API price cuts, creator payouts doubled to $22M, UMG licensing); D3 4→2 (recalibration: private, reinvesting, employee tenders not buybacks, no layoffs; IPO pressure was speculative); D4 5→4 (recalibration: clones proprietary but source audio portable, many competitors); D5 3 unchanged; D6 4→3 (correction: fact audit reduced Qcall 2.8x to one anecdote, free retakes and deletion/opt-out confirmed; no cancellation obstruction documented); D7 5→4 (event: 2026 API, Music and ElevenReader price cuts offset Free-to-Starter feature moves; no ads); D8 3→2 (recalibration: one acqui-hire, licensing-led, no exclusion); D9 3→2 (recalibration: no layoffs, workforce growing, standard VC governance; founder-background speculation dropped); D10 6→4 (recalibration: misuse and training-data litigation show a mixed compliance record, but no lobbying against regulation, legal threats or strategic litigation as 6-7 requires; cooperative with UK/US safety bodies). Eras: 'Beta Launch' re-dated 2023-01-01→2023-01-23 (public beta); 'Unicorn Expansion' re-dated 2024-01-01→2024-01-22 (Series B) and relabeled 'Unicorn and Deepfake Scrutiny'; 'VC Monetization Push' re-dated 2025-01-01→2025-02-28 (voice-data terms update) and relabeled 'Voice-Data Terms Dispute' because the credit-split story rests on retrospective sources; 'IPO Trajectory' re-dated 2026-02-20→2026-02-04 (Series D). All eras re-scored from criteria (15, 22, 31, 30). Trajectory worsening→stable: price cuts and model gains offset tier reshuffles and the new BIPA suit. Alternatives unchanged. Category 'AI' looks right.
Checked all 6 restore candidates against the current fact-audited and regraded record: 0 restored, 0 partly restored, 6 confirmed removed, 0 already present. The surviving core events are present, but none of the dropped sub-claims has adequate evidence for restoration. No content removed and no scores changed. 1. timeline:remove:25 — CONFIRMED REMOVED: Completed December 15, 2025 removal of eleven_monolingual_v1 and eleven_multilingual_v1, resulting broken integrations and lost functionality were not established. The cited December 8 changelog is unrelated (https://elevenlabs.io/docs/changelog/2025/12/8). The September 29 announcement only schedules December removal (https://elevenlabs.io/docs/changelog/2025/9/29); the June 8, 2026 notice instead schedules the same models for July 9 (https://elevenlabs.io/docs/changelog/2026/6/8). Read the current models page and a dated lifecycle tracker (https://vorplabs.com/models/deprecations); neither establishes the originally asserted completed December event or actual customer breakage. The June announcement is already in timeline[42]; it does not restore the dropped completed-removal claim. Three searches, including July 2026 date drift, found no qualifying confirmation. 2. timeline:update:0 — CONFIRMED REMOVED: Exact 12-month stealth-development duration. The original January 23, 2023 launch announcement says founded in 2022, without a stealth duration (https://elevenlabs.io/blog/elevenlabs-raises-2m-pre-seed-and-announces-ai-speech-platform-promising-to-revolutionize-audio-storytelling). Contemporary TechCrunch coverage gives no precise stealth period (https://techcrunch.com/2023/06/20/voice-generating-platform-elevenlabs-raises-19m-launches-detection-tool/). Three exact-phrase, founder-origin and retrospective searches produced only weak retrospective accounts. Read GrowthHunt (https://www.growthhunt.ai/growth-story/elevenlabs), which inconsistently gives twelve, nine and five months; it is not adequate corroboration. Retain the audited launch wording. 3. timeline:update:7 — CONFIRMED REMOVED: February 8 selection date, non-equity support and implied voice-licensing deals. Read the original LinkedIn announcement (publication metadata February 22, 2024), Disney's official February 21 announcement and current program page (https://sites.disney.com/accelerator/2024/02/21/the-walt-disney-company-announces-the-10th-year-of-the-disney-accelerator-program-and-its-2024-participant-companies/; https://sites.disney.com/accelerator/). Disney confirms leadership guidance, which is already preserved, but does not establish non-equity terms or voice-licensing arrangements. Three searches including broader program terms and date drift found no qualifying support. A Crunchbase lead could not be read: live 403 and archive no content. Its snippet was not used as evidence. 4. timeline:update:17 — CONFIRMED REMOVED: First-ever public business-partner departure qualifier and original March 1 announcement date. The original Kukarella page has March 12 publication metadata and says the departure occurred the previous week, without claiming priority (https://www.kukarella.com/post/important-changes-to-our-voice-cloning-technology-transitioning-away-from-elevenlabs). Its August 2025 retrospective also does not establish priority (https://www.kukarella.com/resources/ai-voice-cloning/the-10-best-voice-cloning-tools-in-2025-tested-and-compared). Three searches including the exact qualifier and 2024–2026 coverage found no qualifying confirmation. The departure and stated concerns remain in timeline[20]. 5. timeline:update:20 — CONFIRMED REMOVED: Enterprise complaints as the cause of credit simplification, plus generalized enterprise confusion and opaque-overage assertions beyond the attributed surviving account. Read Flexprice's evolution section (https://flexprice.io/blog/elevenlabs-pricing-breakdown) and Fieldwork (https://fieldwork.news/tools/elevenlabs/). They describe unification and date it to 2025/August 2025 but do not establish enterprise complaints caused it. Three searches checked the exact cause, feedback and August 2025 date drift; no qualifying support was found. Retain the attributed event in timeline[24]. 6. evidence:update:0 — CONFIRMED REMOVED: January 1, 2026 publication date. The cited Qcall page's publication metadata is 2025-06-19T14:17:08+00:00 (https://qcall.ai/elevenlabs-review). Its 2026 title does not establish January publication. Two searches checked the article URL and possible date/title drift; no January 1 date was established. Retain the corrected evidence date. Limitations: All six original-URL Wayback attempts failed with network/429 errors; these failures were not treated as evidence against the claims. All decisions rely on pages actually read in this session and the failure to find qualifying support within the search limits. Search-result snippets were discovery leads only. Tools: 17 search queries in 6 web calls; 30 fetch-page.py invocations, including retries and archives.