Global segment ranking · Artificial intelligence › Generative media
The world’s most valuable voice and audio AI startups
Companies whose main product is AI that generates, understands or transforms voice, speech, music or sound: text-to-speech, voice cloning and conversational voice models, speech recognition and transcription, AI dubbing and voice translation, voice changers and voice enhancement, and AI music and sound generation. Ranked worldwide by AI-estimated valuation, with listed pure players shown in grey at their market value for reference. The market sets the value; AI models estimate it.
Key metrics for this segment: key milestone. Figures as published by each company or the press, dated, not recalculated. The AIs weigh them alongside funding history and comparables.
Not in this segment. Listed, so not ranked: SoundHound AI, Cerence and Neosapience (shown in grey), and Unisound and Mobvoi. Bought or controlled by a group: PlayAI (Meta), Supertone (HYBE), Loudly (Magix), WavTool (Suno), Sureel AI (Warner Music Group); Mureka belongs to Kunlun Tech. Kyutai is a non-profit lab. Fireverse raised money through a token, which Uback never counts. In other segments: voice agents and call-centre AI (Vapi, Retell AI, PolyAI, Smallest.ai) under AI contact centers; dictation and voice companions (Wispr, Sesame AI) under Consumer assistants; podcast tools (Riverside, Wondercraft) under Podcasts; creator audio tools (Descript, LANDR, Splice, Music AI) under Creator tools; video avatars (Synthesia, HeyGen) under Image & video; meeting notes (Otter.ai); game characters (Inworld AI). A company appears in one segment only. Method
The segment at a glance
32companies ranked
2companies on the radar
12countries in the ranking
3listed benchmarks
Sources: Company announcements, TechCrunch, Sifted, tech.eu, Billboard, Music Business Worldwide, Digital Music News, Calcalist, Inc42, 36Kr, Bloomberg, Reuters, and investor announcements. Intention counters will be shown above a threshold of amount and number of Backers.
Rankings by region
Filter the ranking and the Radar. Ranks stay worldwide. Regional pages: Asia · Europe · Middle East & North Africa · North America.Global ranking · Voice & audio
Snapshot27 Oct 2026A snapshot taken on 27 Oct 2026
This ranking shows the startups as they stood on 27 Oct 2026. Funding rounds, valuations and other events announced after that date are not included yet. They will be taken into account in the next edition, scheduled for 13 Nov 2026.
Spotted a more recent round? Tell us →First edition: no movements yet.
This ranking shows the startups as they stood on 27 Oct 2026. Funding rounds, valuations and other events announced after that date are not included yet. They will be taken into account in the next edition, scheduled for 13 Nov 2026.
Spotted a more recent round? Tell us →First edition: no movements yet.
World · 32 ranked companies
| # | Company | Last round · Key metrics | AI valuation | Pool of InterestsInterest: a free, non-binding declaration on a company, or on a sector within an area, made during the free phase of a pool. Pool: all the Interests declared on the same target. This column shows the number of Backers and the combined amount of their tickets, to validated members only. Terms for Backers → | BackBacking a startup on Uback (pronounced “you back”) can be done in two ways: Interest · free: a non-binding declaration that signals your demand for a company or a sector. You can move or withdraw it at any time, with up to 5 at once. Intention · coming soon: a paid, confirmed declaration, still non-binding. Once its pool reaches critical mass, it is presented to a company by a partner desk. Neither is an order nor a commitment to invest. Terms for Backers → |
|---|---|---|---|---|---|
| 1 | ElevenLabs#2 in the United KingdomBuilds AI voice models for text-to-speech, voice cloning, dubbing and conversational agents, sold to creators, developers and enterprises through apps and APIs. | Series D · USD 500M · Feb 2026ARR above USD 330M at end of 2025 · Dec 2025 | Last known equity round: Series D · USD 500M · Feb 2026 (Sequoia Capital, Andreessen Horowitz, ICONIQ) Published valuation: USD 22bn (Sep 2026)Method |
Be the first | Back |
| 2 | Suno#80 in the United StatesGenerates complete songs with vocals and instrumentation from text prompts, for consumers and creators, via a web and mobile app built on proprietary music models. | Series D · USD 400M · Jun 2026USD 300M+ revenue and 2M paid subscribers · Feb 2026 | Last known equity round: Series D · USD 400M · Jun 2026 (Bond Capital; IVP, Forerunner, Union Square Ventures, Alkeon Capital); 'more than USD 400M'; previous round USD 250M Series C at USD 2.45bn post (Nov 2025) Published valuation: USD 5.4bn (Jun 2026)Method |
Be the first | Back |
| — | SoundHound AIListed · Nasdaq · SOUN | USD 1.91bnMarket cap · Oct 2026Listed company · not ranked | |||
| 3 | DeepgramProvides speech-to-text, text-to-speech and voice-agent APIs that developers and enterprises use to transcribe and understand audio in real time. | Series C · USD 130M · Jan 2026More than 1,300 organisations use its models; cash-flow positive in 2025 · Jan 2026 | Last known equity round: Series C · USD 130M · Jan 2026 (AVP; Alkeon Capital, In-Q-Tel, Madrona, Tiger) Published valuation: USD 1.3bn (Jan 2026)Method |
Be the first | Back |
| 4 | VerbitProvides AI-assisted transcription, captioning and audio description for education, legal, media and enterprise customers, combining speech recognition with human editors. | Series E · USD 150M · Nov 2021Key metrics not disclosed | Last known equity round: Series E · USD 150M · Nov 2021 (Third Point Ventures); Series E USD 250M total incl. ~USD 100M secondary; primary ~USD 150MMethod | Be the first | Back |
| 5 | CartesiaDevelops low-latency voice models based on state space models, offering text-to-speech and voice APIs for developers building real-time voice applications. | Series A-II / Series B · USD 100M · Oct 2025Key metrics not disclosed | Last known equity round: Series A-II / Series B · USD 100M · Oct 2025 (Kleiner Perkins; Index Ventures, Lightspeed, NVIDIA)Method | Be the first | Back |
| 6 | GradiumDevelops low-latency voice models for speech recognition, speech synthesis and real-time voice agents, spun out of the Paris research lab Kyutai. | Seed (incl. extension) · USD 100M · Dec 2025Key metrics not disclosed | Last known equity round: Seed (incl. extension) · USD 100M · Dec 2025 (FirstMark Capital, Eurazeo; NVIDIA, Xavier Niel, Rodolphe Saadé, Eric Schmidt); USD 70M (Dec 2025) + USD 30M extension (Jul 2026) added to the same roundMethod | Be the first | Back |
| — | CerenceListed · Nasdaq · CRNC | USD 380MMarket cap · Oct 2026Listed company · not ranked | |||
| 7 | SpeechmaticsProvides speech recognition and voice AI models in many languages, sold as cloud APIs and on-premise software to enterprises and developers. | Series B · USD 62M · Jun 2022Revenue about USD 18.7M in 2025 (Sacra estimate) · 2025 | Last known equity round: Series B · USD 62M · Jun 2022 (Susquehanna Growth Equity; AlbionVC, IQ Capital)Method | Be the first | Back |
| 8 | AISpeechDevelops speech recognition, voice interaction and dialogue AI for smart cars, consumer electronics and enterprise customers in China. | Series E · USD 58M · Apr 2020Key metrics not disclosed | Last known equity round: Series E · USD 58M · Apr 2020Method | Not open to BackersCompanies based in mainland China are ranked like any other startup, but Uback does not open pools on them: capital controls and offshore (VIE) structures make minority deals impractical for our Backers. You can still declare an interest in this segment → |
|
| 9 | Fish AudioOffers expressive text-to-speech and voice cloning models through a consumer platform and developer API used by creators and app makers. | Seed · USD 52M · Jul 2026ARR of about USD 21M and over 8 million users · Jul 2026 | Last known equity round: Seed · USD 52M · Jul 2026 (Coreline Ventures, Capital Today; 359 Capital, Play Time, HF0, 645 Ventures)Method | Be the first | Back |
| 10 | Hume AIDevelops emotionally expressive voice and speech-to-speech models and an empathic voice interface API that developers license to build voice products. | Series B · USD 50M · Mar 2024CEO says on track for USD 100M revenue in 2026 (company claim) · Jan 2026 | Last known equity round: Series B · USD 50M · Mar 2024 (EQT Ventures; Union Square Ventures, Comcast Ventures, LG Technology Ventures)Method | Be the first | Back |
| 11 | AssemblyAIOffers speech recognition and audio intelligence models through an API that developers use to transcribe, summarise and analyse voice data. | Series C · USD 50M · Dec 2023Key metrics not disclosed | Last known equity round: Series C · USD 50M · Dec 2023 (Accel; Smith Point Capital, Insight Partners, Y Combinator)Method | Be the first | Back |
| 12 | ACE StudioOffers AI virtual singers and a music creation workstation where producers enter melody and lyrics, plus a consumer song app and open-source ACE-Step models. | Series B · USD 40M · Sep 2026Nearly 100,000 paying creators; monthly revenue in USD millions · Sep 2026 | Last known equity round: Series B · USD 40M · Sep 2026 (CCV, Shunwei Capital, Alphaist; existing Wuyuan Capital, Boyu Capital, HF0); reported as 'nearly USD 40M'Method | Not open to BackersCompanies based in mainland China are ranked like any other startup, but Uback does not open pools on them: capital controls and offshore (VIE) structures make minority deals impractical for our Backers. You can still declare an interest in this segment → |
|
| 13 | RimeBuilds text-to-speech and speech-to-speech models designed for enterprise phone calls, used by contact centres and healthcare providers for natural-sounding voices. | Series A · USD 24M · Jul 2026Powers nearly 100 million phone calls per month · Jul 2026 | Last known equity round: Series A · USD 24M · Jul 2026 (M13; Twilio Ventures, Corazon Capital, Unusual Ventures)Method | Be the first | Back |
| 14 | DeepdubProvides AI dubbing and voice localisation for film, television and games, generating translated speech that preserves the original voice. | Series A · USD 20M · Feb 2022Key metrics not disclosed | Last known equity round: Series A · USD 20M · Feb 2022 (Insight Partners)Method | Be the first | Back |
| 15 | GladiaProvides a real-time multilingual speech-to-text and audio intelligence API that software companies embed in meeting, call and voice-agent products. | Series A · USD 16M · Oct 2024Key metrics not disclosed | Last known equity round: Series A · USD 16M · Oct 2024 (XAnge)Method | Be the first | Back |
| 16 | EndelGenerates personalised functional soundscapes for focus, relaxation and sleep using AI that adapts to time, weather and biometric data, through apps and partnerships. | Series B · USD 15M · Apr 2022Key metrics not disclosed | Last known equity round: Series B · USD 15M · Apr 2022 (Waverley Capital, True Ventures)Method | Be the first | Back |
| 17 | VoicemodOffers real-time voice changer and soundboard software and AI voices for gamers, streamers and creators on desktop and mobile. | Series B · USD 14.5M · Nov 2022Key metrics not disclosed | Last known equity round: Series B · USD 14.5M · Nov 2022; EUR 14.5M per El Referente (≈ USD 14.5M at late-2022 parity); announced with acquisition of Voctro LabsMethod | Be the first | Back |
| 18 | AudioShakeSeparates recordings into stems such as vocals, dialogue, music and effects using AI models, for labels, publishers, broadcasters and developers via API and apps. | Series A · USD 14M · Oct 202540+ enterprise clients incl. Universal, Warner and BMG · Oct 2025 | Last known equity round: Series A · USD 14M · Oct 2025 (Shine Capital; Thomson Reuters Ventures, Origin Ventures, Background Capital, Indicator Ventures)Method | Be the first | Back |
| 19 | Resemble AIOffers voice cloning and speech synthesis alongside deepfake detection models that enterprises and security platforms use to generate and verify synthetic media. | Strategic round · USD 13M · Dec 2025Key metrics not disclosed | Last known equity round: Strategic round · USD 13M · Dec 2025 (Google AI Future Fund, Okta Ventures, Taiwania Capital, Gentree Fund)Method | Be the first | Back |
| — | NeosapienceListed · KOSDAQ · 0161M0 | USD 67MMarket cap · Oct 2026Listed company · not ranked | |||
| 20 | WellSaid LabsProvides studio-quality AI text-to-speech voices that enterprise teams use to produce narration for training, marketing and product content. | Series A · USD 10M · Jun 2021Key metrics not disclosed | Last known equity round: Series A · USD 10M · Jun 2021 (FUSE; Voyager Capital, Qualcomm Ventures, Good Friends)Method | Be the first | Back |
| 21 | Murf AIOffers an AI voice generator and text-to-speech API with studio editing tools, used by creators and businesses to produce voiceovers in many languages. | Series A · USD 10M · Sep 2022Key metrics not disclosed | Last known equity round: Series A · USD 10M · Sep 2022 (Z47; Elevation Capital)Method | Be the first | Back |
| 22 | UdioGenerates songs from text prompts and is rebuilding its service as a licensed platform where fans remix, cover and create music with label-cleared models. | Seed · USD 10M · Apr 2024Licensing deals with UMG, Warner, Kobalt, Merlin, Believe and NMPA · Jul 2026 | Last known equity round: Seed · USD 10M · Apr 2024 (Andreessen Horowitz)Method | Be the first | Back |
| 23 | Gaudio LabDevelops AI audio technology including spatial audio, loudness normalisation, stem separation and sound generation, licensed as SDKs to streaming, device and media companies. | Series B · USD 9.5M · Oct 2021Key metrics not disclosed | Last known equity round: Series B · USD 9.5M · Oct 2021 (LB Investment; Samsung Venture Investment, Naver D2 Startup Factory, DS Asset Management, Capstone Partners); KRW 11.3bn converted at ~1,185 KRW/USD; cumulative KRW 16.9bnMethod | Be the first | Back |
| 24 | KrispProvides AI noise cancellation, accent conversion and voice processing software for calls, used by remote workers and contact centres. | Series A extension · USD 9M · Feb 2021Key metrics not disclosed | Last known equity round: Series A extension · USD 9M · Feb 2021Method | Be the first | Back |
| 25 | LOVOOffers an AI voice generator and video editor with text-to-speech and voice cloning, used by marketers, educators and creators to produce voiceovers. | Pre-Series A · USD 4.5M · Aug 2021Key metrics not disclosed | Last known equity round: Pre-Series A · USD 4.5M · Aug 2021Method | Be the first | Back |
| 26 | CAMB.AIProvides AI dubbing, voice translation and speech models that broadcasters, sports leagues and studios use to localise audio and video into many languages. | Seed · USD 4M · Feb 2024Key metrics not disclosed | Last known equity round: Seed · USD 4M · Feb 2024 (Courtside Ventures)Method | Be the first | Back |
| 27 | Producer.aiGenerates and edits music through a conversational AI producer that composes, arranges and remixes tracks from prompts, formerly known as the Riffusion app. | Seed · USD 4M · Oct 2023Key metrics not disclosed | Last known equity round: Seed · USD 4M · Oct 2023 (Greycroft; Spark Capital, Sky9 Capital, South Park Commons)Method | Be the first | Back |
| 28 | SoundfulGenerates royalty-free music tracks and stems in chosen genres for content creators, brands and producers through an AI platform with licensing included. | Seed · USD 3.8M · Apr 2022Key metrics not disclosed | Last known equity round: Seed · USD 3.8M · Apr 2022 (Disney, Universal Music Group, Accenture, Beatport)Method | Be the first | Back |
| 29 | BoomyLets users create instant songs with AI and distribute them to streaming services, sharing royalties, through a consumer web platform. | Seed · USD 3.1M · Jun 2022Key metrics not disclosed | Last known equity round: Seed · USD 3.1M · Jun 2022 (Warner Music Group, Scrum Ventures, Boost VC, First Check Ventures); USD 3.1M is total raised per VCBacked; split by round unclearMethod | Be the first | Back |
| 30 | SOUNDRAWGenerates royalty-free background music that creators customise by genre, mood and length, with a B2B API for platforms and backing tracks for artists. | Undisclosed · USD 3M · Mar 2024Key metrics not disclosed | Last known equity round: Undisclosed · USD 3M · Mar 2024 (Carbide Ventures; mint VC, Ceres, iSGS, SMBC Venture Capital)Method | Be the first | Back |
| 31 | Beatoven.aiGenerates royalty-free background music tailored to mood and video length for content creators and businesses, via a web app and API. | Pre-Series A · USD 1.3M · Mar 2024Over 1 million users; 96% of revenue international · Mar 2024 | Last known equity round: Pre-Series A · USD 1.3M · Mar 2024 (Capital 2B; Entrepreneur First, IvyCap Ventures, Upsparks Capital, Rukam Capital); INR 11 crore at ~83 INR/USDMethod | Be the first | Back |
| 32 | Respeecher#17 in UkraineProvides speech-to-speech voice conversion and cloning used by film, television, games and healthcare to reproduce voices with consent. | Pre-Series A · USD 1M · Dec 2023Key metrics not disclosed | Last known equity round: Pre-Series A · USD 1M · Dec 2023 (ff Venture Capital)Method | Be the first | Back |
Back · Backing on Uback (pronounced “you back”) means declaring an Interest (free) or an Intention (coming soon). Backing does not mean that this company is raising funds or that a shareholder is selling shares, and it is not a commitment to invest. Any deal, through a funding round or a sale by an existing shareholder, is handled by a partner desk: the Fundraising Desk of the country for a round, the Secondary Sales Desk assigned to the company for a sale. · Terms
Corporate investor? Back the startups of your own segment
No ranked company in this region yet: see the Radar.
Ranking in descending order of AI-estimated valuation, based on public information, with no human intervention on the order. The valuations, ranges and brackets shown are editorial estimates produced by AI, purely indicative and imperfect by nature: neither a financial valuation, nor an offer, nor investment advice. Only the market sets a company’s value, through a funding round or a sale. Ranking established by Claude (Anthropic) from public information, following the published rules. A consensus of several AIs will apply from 1 December 2026. Eligibility: at least $1M raised in equity, non-listed company, main activity in this segment (voice and audio AI). Country shown is the country of main operations. Request a correction · Full method →
Get every new ranking
New entries, exits and movements, at every edition.
No data passed on to third parties. See our Privacy policy.
Challengers
A registered company raising funds that is not in the ranking will be able, for a fee, to appear in a separate list labelled “Challengers”, for a set period, with a memo available to verified Backers. The listing has no effect on the ranking. Not open yet.
Sponsored space · no Challenger listed yet
Radar · known eligible companies, not ranked
Sorted by date of last round, with no judgement and no AI call.
- Pozalabs · South Korea · AsiaOct 2022Back
- Speechify · United States · North AmericaJan 2021Back
No company on the radar in this region yet.
Listed benchmarks
Not ranked: no intention is possible on a listed company.
- SoundHound AI · United StatesNasdaq · SOUN · USD 1.91bn (Oct 2026)
- Cerence · United StatesNasdaq · CRNC · USD 380M (Oct 2026)
- Neosapience · South KoreaKOSDAQ · 0161M0 · USD 67M (Oct 2026)
Sources and method for this segment
Calendar. Global segment rankings are published twice a year. Next edition: November 13, 2026.
Eligibility. At least $1M raised in equity, non-listed company, main activity in this segment (voice and audio AI). A company appears in one segment only.
Sources. Company announcements, TechCrunch, Sifted, tech.eu, Billboard, Music Business Worldwide, Digital Music News, Calcalist, Inc42, 36Kr, Bloomberg, Reuters, and investor announcements.