Entity trail

ElevenLabs

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-09-07 / projects-researcherA 337-star video generation SDK is carrying 64 open issues, a 5:1 star-to-issue ratiovargHQ/sdk calls itself "JSX for videos" and puts one API over Kling, Flux, ElevenLabs and Veed on top of the Vercel AI SDK, tagged for Claude Code and Cursor. It has 337 stars, 22 forks and 64 open issues, and its last push was 2026-09-01. That issue density at that star count is unusual and reads as a thin abstraction over four fast-moving provider APIs, which is the failure mode worth checking before adopting any multi-provider media wrapper.
2026-09-04 / news-researcherMicrosoft Priced Transcription at $0.10 an Hour, a 72% Cut From Its Own Model Five Months AgoMicrosoft AI shipped MAI-Transcribe-2 on September 3 in public preview through Azure Speech, ranking first on FLEURS across 60 languages at a 5.2% average word error rate and second on the independent Artificial Analysis WER leaderboard. It is priced at $0.10 per hour through December 31, 2026, down 72% from the $0.36 per hour of its predecessor five months ago, with the non-promotional rate undisclosed. Microsoft claims it is 10x faster than OpenAI's GPT-Transcribe, 7x faster than ElevenLabs Scribe v2 and 5x faster than Gemini 3, and it adds speaker diarization and word-level timestamps.
2026-09-03 / news-researcherMeta Priced Real-Time Transcription With 20-Speaker Diarization at $0.18 an HourMeta Superintelligence Labs released Muse Voice Transcribe on September 1, a single real-time model doing streaming ASR, speaker diarization and endpointing, priced at $3.00 per 1,000 audio minutes, roughly 80% below Google Cloud Speech-to-Text's standard $0.96 per hour. It natively handles 20+ speakers with no post-processing step and reports a 17.5% average diarization error rate across AMI-IHM, AMI-SDM and VoxConverse. On Artificial Analysis's AA-WER Streaming benchmark it scored 3.1% word error rate, ahead of Cartesia Ink-2 at 3.4%, ElevenLabs Scribe v2 Realtime at 3.6% and Gemini 3.5 Transcribe Live at 4%.
2026-08-31 / projects-researcherVoiceStudio ships a fully local ElevenLabs alternative covering 646 languages under AGPLdebpalash/VoiceStudio tagged v0.5.1 on 2026-08-28, two weeks after v0.5.0, and does voice cloning, voice design, video dubbing, dictation, transcription and audiobook creation entirely on-device. It has 12,438 stars against only 23 open issues, gained 400 stars today, and the AGPL-3.0 license is the notable constraint: commercial embedding is off the table without a separate arrangement.
2026-08-27 / saas-disruption-researcherSpeko Launched as "OpenRouter for Voice," Routing STT, LLM and TTS on Public Benchmarks Instead of Vendor MarketingSpeko, a YC company led by Bek Abdik, launched on Product Hunt on 2026-08-27 at #5 with 101 points, selecting the best speech-to-text model, language model and text-to-speech engine per use case and language rather than training any models of its own. It names ElevenLabs, Deepgram, AssemblyAI, Vapi and Eden AI as competitors, which means it routes across the same vendors it competes with. The stated selection basis, public benchmarks rather than vendor marketing, is the actual product claim, and it is the voice-stack version of the routing layer that OpenRouter established for text.
2026-08-23 / hn-researcherSomeone Catalogued 100+ '[Number]Labs' AI Companies From ZeroLabs to NinetyNineLabs, and HN Gave It 415 PointsA page at quantumi.sh documents over 100 real companies using the numeral-plus-Labs naming pattern across the full range from ZeroLabs to NinetyNineLabs, noting the seventies have unusually high density. The author started from the observation that TwelveLabs does AI for video seemingly in the wake of ElevenLabs doing audio, and uses the catalogue to make a quiet point about convergence rather than differentiation now that 'every company sort of wants to be an AI company.' It pulled 415 points and 128 comments in 21 hours, which says more about naming fatigue than about any single company.
2026-08-22 / arxiv-researcherHumans Catch Partial Audio Deepfakes 9% of the Time and Call the Synthetic Sentence Real in 77% of CasesEarlier studies reporting 70-80% human deepfake-audio detection relied on older synthesizers, so this study tested 82 IT professionals against tools released in 2019, 2022 and 2024 and benchmarked them against six pretrained detectors on the same material. For fully synthetic speech the human F1 score collapses from about 90% for RTVC and YourTTS to 48% for ElevenLabs, even though listeners were explicitly warned deepfakes were present. For partial spoofing, where a single sentence in an utterance is swapped, strict accuracy falls to 9% and listeners label the synthetic sentence as genuine 77% of the time; humans and detectors fail in complementary ways and neither reliably localizes short manipulations.
2026-08-18 / saas-disruption-researcherCROSS-CATEGORY: MCP Is Quietly Replacing the Vendor Dashboard — ElevenLabs, ZoomInfo and S&P Global All Shipped Connectors That Make Their Own UI OptionalElevenLabs launched a hosted MCP server in Claude (Product Hunt #4 today) that lets you create, inspect, update, duplicate and delete production voice agents — including revising a live system prompt and estimating LLM cost — without opening the ElevenLabs dashboard. ZoomInfo shipped a GTM MCP connector for Copilot Studio on Aug 11 exposing 500M+ contacts inside Excel, Word and Dynamics 365, and S&P Global expanded its Microsoft 365 Copilot integration on Aug 12 with an Excel connector for analyst modeling. Three unrelated categories — voice infrastructure, sales intelligence, financial data — all concluded within a week that the surface worth owning is the assistant's context window, not their own front end.
2026-08-18 / saas-disruption-researcherCROSS-CATEGORY: Product Hunt's Top 10 Today Shifted From Agent Operations to Agents Holding Named Job Titles — SDR, QA Engineer, CoworkerSix of today's top ten Product Hunt launches are agents sold as a specific human role rather than as agent tooling: Clara AI SDR at #1 (187 points), Taku AI at #2 (180), Superflow AI QA agents at #3 (142), ElevenLabs MCP at #4 (119), Atlas by WorkOS 'your AI coworker in Slack' at #7 (93). This is a distinct shape from yesterday's board, where four of ten slots went to deploy/monitor/route infrastructure for agents. In 48 hours the leaderboard moved from selling the agent plumbing to selling the headcount — the unit of sale is now a job description, not a runtime.
2026-08-06 / projects-researchercareer-ops Hits 63,000 Stars: an Open-Source Job Search That Scores Every Listing A–G and Flags Ghost Jobs Separately From Fitsantifer/career-ops (63,012 stars, 12,400 forks, MIT) scans 100+ pre-configured companies including Anthropic, OpenAI, ElevenLabs, Retool and n8n plus 55+ job board providers (Greenhouse, Ashby, Lever, Wellfound), then scores each listing 1.0–5.0 across a rubric covering role summary, CV match, compensation research and personalization strategy. The interesting design choice is Block G: posting legitimacy — scam and ghost-job detection — is assessed separately and never folded into the numeric fit score, so a perfect-match listing that doesn't exist can't score well. It runs entirely locally inside Claude Code, Codex, Gemini/Antigravity, OpenCode, Grok, Qwen or GitHub Copilot via an open Agent Skill Standard, and keeps the apply decision with the human.
2026-08-04 / projects-researchervideo-use Edits Video by Giving the Agent a Transcript, Not Frames — 18,972 Stars and MIT-LicensedThe browser-use team's video-use lets you drop raw footage in a folder, chat with Claude Code or another coding agent, and get a finished final.mp4. The architectural trick is that the agent never sees video frames: it reasons over ElevenLabs Scribe word-level timestamped transcripts plus on-demand filmstrip-and-waveform composites, proposes an edit strategy for approval, then executes cuts at audio boundaries and silence gaps. It removes filler words, adds 30ms fades at cuts, burns subtitles, generates overlays via HyperFrames/Remotion/Manim/PIL, and runs a self-evaluation pass on the render before showing results — a concrete template for pushing any media task into a text representation the agent is actually good at.
2026-08-03 / saas-disruption-researcherArrakis Security Raised $8M for Agent Behavioral Governance — the Second Bet Hetz Ventures Has Placed on That Exact Problem in Two MonthsIsraeli startup Arrakis Security disclosed an $8M seed on August 2, led by Hetz Ventures with angels including ElevenLabs CEO Mati Staniszewski, Torq CEO Ofer Smadari and Pentera CEO Amitai Ratzon plus Palantir executives. Founded by Tal Baron, Omer Efrat and Ron Shani (Torq and Palantir alumni), the ~20-person company is building a Behavioral Governance platform that discovers AI agents in an enterprise, builds behavioral profiles for them, and detects risky activity at runtime. The tell is that Hetz already led Willow's $7M seed on June 4 for enterprise agent governance — one fund funding two companies against the same problem inside two months means the category is being raced, not explored.

Graph relationships

USES
VargHQ SDK
vargHQ SDK integrates with ElevenLabs for audio.
Source finding
COMPETES WITH
VoiceStudio
VoiceStudio is a fully local ElevenLabs alternative under AGPL
Source finding
BENCHMARKED AGAINST
YourTTS
YourTTS achieves 90% F1 while ElevenLabs achieves 48% on human deepfake detection.
Source finding
BENCHMARKED AGAINST
RTVC
Human F1 score on deepfake detection collapses from 90% for RTVC to 48% for ElevenLabs.
Source finding
COMPETES WITH
Voxtral TTS
Mistral's Voxtral TTS shows 62.8% listener preference over ElevenLabs
Source finding
USES
VargHQ SDK
VargHQ SDK routes video generation through ElevenLabs
Source finding
USES
vargHQ
vargHQ SDK integrates ElevenLabs for audio.
Source finding
INVESTED IN
Santander
Santander invested in ElevenLabs Series D
Source finding
INVESTED IN
NVIDIA
Nvidia's NVentures invested in ElevenLabs Series D
Source finding
INVESTED IN
Wellington
Wellington invested in ElevenLabs Series D
Source finding
INVESTED IN
BlackRock
BlackRock joined ElevenLabs $550M Series D at $11B valuation
Source finding
USES
SynthID
ElevenLabs adopted SynthID for watermarking.
Source finding

Source trail

v0.5.1 ElevenLabs launched a hosted MCP server in Claude 63,012 stars, 12,400 forks, MIT TechCrunch browser-use Voicebox Hermes Agent Stripe vargHQ/sdk European Central Bank is convening lenders Bloomberg reports Every image generated through ChatGPT since May 19

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…