videoengine.ai

Issue 25 · 19 Nov 2026 · A weekly research record of AI video

Market

Revenue, funding and share figures, research notes on China and tool benchmarks, a ledger of 81 dated claims, interview excerpts and a landscape of 59 tools.

MethodClaims come from SOPHIA, our industry knowledge base: one atomic claim per row with its source, the date the fact was true, a source tier (A primary or top press, B reputable secondary) and a confidence score. Only claims with a public source are published. Company-reported figures are labeled as such. LimitationMost claims come from tier B secondary reporting; where a primary source exists it is cited instead.

Findings

  1. Higgsfield reported annualized revenue of US$700 million in August 2026, up from about US$200 million at the end of 2025, and raised a US$400 million Series B at a US$5.4 billion valuation. Both figures are company-reported; the company has since claimed higher run rates, listed in its company file.1
  2. Kling AI's annualized revenue run rate was about US$500 million in March 2026, four times the level a year earlier, and about 75% of its revenue comes from outside China, according to Bloomberg via KR Asia. In July 2026 it raised a reported US$2.8 billion to nearly US$3 billion at a US$18 billion post-money valuation.2,3
  3. ByteDance's Seedance held more than 80% of China's AI video market measured by daily token or compute consumption, and reached 95% of the short-drama industry, according to 36Kr reports from July and September 2026 (tier B). A separate estimate in our China research notes, 80 to 90% of AI-drama workflows, measures usage rather than compute.4,5
  4. OpenAI closed the Sora consumer app on 26 April 2026. Its total lifetime revenue was about US$2.1 million against an operating cost of about US$1 million a day, according to TechTimes (tier B).3
  5. MiniMax, maker of Hailuo, reported first-half 2026 revenue of US$116.6 million, up 283% year on year; revenue from its AI-native products, Hailuo AI among them, grew 101% to US$42.6 million. Both figures are company-reported.6
  6. Aishi Technology, maker of PixVerse, closed about US$440 million of Series C funding by July 2026 at a valuation above US$2 billion. Runway raised US$315 million at a US$5.3 billion valuation in February 2026.7,8

Notes

  1. 1PR Newswire, Higgsfield Raises $400 Million Series B at $5.4 Billion Valuation with Annualized Revenue Reaching $700 Million, 17 Aug 2026. Tier A source. SOPHIA claims 609, 608.
  2. 2KR Asia / Bloomberg, Kuaishou's Kling AI seeks fresh funding ahead of planned 2027 IPO, 1 Jul 2026. Tier B source. SOPHIA claims 585, 586, 584.
  3. 3TechTimes, Kling AI Raises Record $2.8 Billion: What Creators Owe China Under Beijing Law, 4 Jul 2026. Tier B source. SOPHIA claims 1, 6.
  4. 436Kr English, AI Video Generation Showdown in July: Over 200 Billion Yuan of Capital Pours Into the Track, Who Has Made It to the Final Round?, 28 Jul 2026. Tier B source. SOPHIA claims 7, 10.
  5. 536Kr English, MiniMax & Alibaba Simultaneously Launch Video Agent Businesses: Post-Model Competition, Is Workflow the Next Battleground?, 3 Sep 2026. Tier B source. SOPHIA claim 19.
  6. 6MiniMax (Company IR), MiniMax Announces First Half 2026 Financial Results, 1 Aug 2026. Tier A source. SOPHIA claim 605.
  7. 7TechCrunch, Video-generation startup PixVerse raises $439M, valuation soars past $2B, 13 Jul 2026. Tier A source. SOPHIA claim 617.
  8. 8TechCrunch, AI video startup Runway raises $315M at $5.3B valuation, eyes more capable world models, 10 Feb 2026. Tier A source. SOPHIA claims 589, 590.
  9. 9Morphed.app / Fortune Business Insights / Grand View Research, AI Video Generation Statistics 2026: Market Size & Adoption, 1 Sep 2026. Tier B source. SOPHIA claim 635.

Research notes

China AI video

Hongguo top 10 had zero live-action dramas for the first time

As of 17 Oct 2026 · 4 sources

AI-made titles have become the default supply in Chinese micro-drama. On 5 October 2026 the Hongguo chart top 10 was led by AI anime-style drama, with no live-action title present. Chinese trade press reports that the large majority of new micro-dramas are now AI-made.

  • On 2026-10-05 the Hongguo top 10 contained no live-action dramas for the first time, led by AI anime-style drama, per Sina. k.sina.com.cn ↗
  • More than 430,000 micro-dramas went online from January to August 2026, over 90% AI-made, roughly 1,800 a day, per Huxiu. huxiu.com ↗
  • Seedance is reportedly used in 80% to 90% of AI-drama workflows, per 21jingji. 21jingji.com ↗
  • In April 2026 an AI version of a drama became the first AI drama to beat a live-action drama on Hongguo, per 21jingji. m.21jingji.com ↗

China AI short drama projected near RMB 40B in 2026

As of 4 Nov 2026 · 4 sources

Chinese sources put AI short-drama revenue above RMB 22B in the first half of 2026, with about RMB 40B projected for the full year. The wider Chinese micro-drama market already exceeds the domestic box office. Success is highly concentrated: fewer than 1% of titles reach 100M plays.

  • AI short dramas generated more than RMB 22B in H1 2026, with about RMB 40B projected for 2026 and 600M users, per 21jingji. 21jingji.com ↗
  • China's micro-drama market was RMB 50.44B in 2024, larger than the China box office, per a CNSA white paper reported by Yicai. yicai.com ↗
  • China's micro-drama market reached RMB 67.79B in 2025, per CNSA via Xinhua. news.cn ↗
  • Of about 221,900 AI titles on Douyin in H1 2026, only 1,055 (0.47%) passed 100M plays, per Huxiu citing DataEye. huxiu.com ↗

Platform payouts on Chinese AI dramas fell about 90% in 2026

As of 4 Nov 2026 · 3 sources

Revenue share paid by Hongguo and Douyin on free AI dramas dropped sharply through 2026 as guarantees were removed and coefficients cut. Creator-reported payouts are now in the single digits of RMB per 10,000 plays. Combined with a sub-1% hit rate, platform share is described by creators as a lottery for individuals.

  • Hongguo's share formula multiplies valid watch time by a per-minute rate, a genre coefficient and a copyright coefficient; coefficients changed three times in six months, per The Paper. m.thepaper.cn ↗
  • One creator reported RMB 666 earned on 1.64M plays, about RMB 4 per 10,000 plays. bilibili.com ↗

Chinese AI drama compute costs roughly RMB 150 to 800 per finished minute

As of 4 Nov 2026 · 3 sources

Per-second generation prices in China are low, but creators report that only a small share of takes is usable, which multiplies real cost. Studio all-in rates sit near RMB 1,000 per minute, with labour the larger share. A solo creator can produce a 50-episode drama in about a week.

  • A 50-episode AI drama took one person about 7 days and about RMB 20k (US$2.8k) of compute, per a Bilibili industry breakdown. bilibili.com ↗
  • Only 4% to 10% of generated footage is usable, per one September 2026 report, putting real compute at roughly RMB 150 to 800 per finished minute. post.smzdm.com ↗

Client work, not platform share, carries Chinese AI video income

As of 4 Nov 2026 · 4 sources

A study of 28 Bilibili creator videos found the steadiest AI video income in China comes from client production, brand posts and commerce rather than platform revenue share. Subcontract rates are falling quarter by quarter. Income claims from course sellers could not be verified.

  • A 1-minute AI brand film prices at about RMB 22k (US$3.1k) against about RMB 2,000 of compute and 40 hours of work. bilibili.com ↗
  • Subcontract manju production pays up to RMB 2,500 per minute at the top, RMB 1,000 to 2,000 at market rate, and RMB 300 to 500 for low-motion work. bilibili.com ↗
  • AI tool companies pay creators RMB 20k to 40k per sponsored post at 10k to 30k followers via agencies; one real MiniMax deal paid RMB 11,500 less a RMB 2,500 agency rebate. bilibili.com ↗
  • Douyin sellers report single AI ads moving RMB 100k to 200k of goods. bilibili.com ↗
RouteWho paysTypical rate (RMB)
Platform share on free AI dramasPlatform, from ad revenue5 to 10 per 10k plays
Subcontract productionDrama producers, platforms1,000 to 2,500 per minute
Client work: brand films, adsBrands, SMEs, agenciesabout 22k per 1-min brand film
Sponsored creator postsAI tool companies via agencies20k to 40k per post
Affiliate commerceMerchants, by commissionabout 3 to 20 per sale
Novel promotion videosNovel apps, per new readerabout 9 per new user

China requires AI labels and filing for every AI micro-drama

As of 4 Nov 2026 · 3 sources

China has layered AI content labelling, mandatory filing and budget-tiered review onto AI micro-drama since 2025. A new NRTA micro-drama management rule took effect on 1 September 2026. Platforms have also begun removing near-identical AI titles.

  • AI content labelling rules from CAC, MIIT, MPS and NRTA, requiring visible and hidden metadata labels, took effect on 1 September 2025. cac.gov.cn ↗
  • From 1 July 2026 AI micro-dramas are classified by budget: RMB 800k and above or sensitive themes as key titles, RMB 300k to 800k as ordinary, and under RMB 300k reviewed by the platform, per Huxiu. huxiu.com ↗
  • Hongguo removed about 10,000 non-compliant AI anime dramas in July 2026 for near-identical covers and blurbs, per Sohu. sohu.com ↗

Course sellers make the loudest AI video income claims in China

As of 4 Nov 2026 · 3 sources

A large share of popular Chinese AI video tutorials end in paid-course funnels, and the biggest income claims come from those sellers. Consumer complaints about AI courses have risen. Skeptical creators publish the hardest loss figures.

  • A skeptic creator argued the industry lacks traffic, merchant resources and order channels, not people who can make AI dramas. bilibili.com ↗

Stylised AI dramas lead Chinese charts while photoreal faces draw backlash

As of 17 Oct 2026 · 6 sources

The leading Chinese AI dramas of August to October 2026 were mostly stylised fantasy and time-slip series released in fast multi-season runs. One hit IP is now being remade in live action. Photoreal AI faces in long-form and licensed-likeness projects drew public criticism.

  • Record of Ten Thousand Demons passed about 6B plays over 12+ seasons; its studio started with 6 people, and a live-action version began shooting on 2026-09-20, per QQ News. news.qq.com ↗
  • Best Actress Yaya, launch title of Tencent's free ad-supported Green Bubble drama app in WeChat, drew hundreds of millions of plays in 3 days and a 2M concurrent peak, per Sina. k.sina.com.cn ↗
  • Mango TV's AI series Post-Journey to the West won its provincial prime-time slot but drew a 'dead face' backlash, per Huxiu. huxiu.com ↗
  • An iQiyi AI drama using licensed actor likenesses triggered a trending 'creepy' hashtag, per IT Home. ithome.com ↗
Title (English)PlatformReported metricGenre
Little Nuonuo Comes Down the MountainHongguoS2 over 100M plays day one; S3 #1 at 86.2M heatCultivation fantasy
Record of Ten Thousand DemonsHongguoAbout 6B plays over 12+ seasonsSupernatural demon hunter
Who Says a Playboy Can't Top the ExamDouyin, Hongguo1.78B Douyin plays, 788 episodesTime-slip
Best Actress YayaTencent Green Bubble (WeChat)Over 200M heat; 2M concurrent peakAI photoreal time-slip
Three Days Until I Time-Travel, So I Emptied the MallHongguo65.0M heatTime-slip, stockpiling
In-laws Changed the Locks, So I Took Back the Marital HomeHongguo AI photoreal chart59.5M heat, #1 in SeptemberFamily revenge

Tool benchmarks

Chinese video models range from RMB 0.18 to 1.38 per second

As of 4 Nov 2026 · 3 sources

A second-hand comparison of effective per-second costs across Chinese AI video tools, drawn from media price tests rather than official price pages. Aggregator access to Kling via LibLib was the cheapest found; Jimeng VIP access to Seedance 2.0 was the most expensive at 720p. Credits per Jimeng clip reportedly rose about 10x during 2026.

  • AI Xinbang's price test, published on Tencent News, compared effective per-second costs across major Chinese video tools. news.qq.com ↗
  • Seedance 2.0 API pricing on Volcano Engine is RMB 46 per million tokens without video input and RMB 28 with video input, about RMB 1 per second, per IT Home. ithome.com ↗
  • Jimeng peak queues reached about 13,000 people and 11-hour waits, per Cailian Press. cls.cn ↗
ToolAccess or planEffective RMB per secondNote
Seedance 2.0 (Jimeng)Premium VIP, 720p1.38Most expensive found
Seedance 2.0 APIVolcano Engineabout 1.0About RMB 15 per 15s clip
Hailuo 2.31080p, annual plan0.93
PixVerse V6Elite plan0.48
Kling 3.0 OmniBlack Gold annual, 720p0.43
ViduAnnual plan0.42
OiiOiiAPEX plan0.37One-stop manju platform
Seedance 2.5480p0.31080p text-to-video about 5.61
Wan 3.0 (via Qwen)720punder 0.5Vendor-sponsored figure
Kling 3.0 via LibLibAnnual plan0.18Cheapest found

Market size and economics

Advertising, not gaming, is the largest pool for AI video

As of 17 Oct 2026 · 5 sources

Global games revenue of about US$202B in 2025 is roughly six times global box office and about equal to cinema plus streaming combined. Global advertising, at about US$1.0T to US$1.14T in 2025, is five to six times larger than gaming. The bulk of video spend relevant to AI video sits in advertising creative.

  • Global games revenue reached US$201.6B in 2025 (+9.1%), with mobile at US$113.3B, per Newzoo via PC Gamer. pcgamer.com ↗
  • Global box office was about US$33.6B in 2025 on Gower Street's year-end estimate. screendaily.com ↗
  • PwC sized OTT streaming video at US$169B in 2024, rising to US$230B by 2029. pwc.com ↗
  • Global advertising reached US$1.14T in 2025 per WPP Media's December update. marketingbrew.com ↗
  • US digital video ad spend is forecast at US$81.9B in 2026, including CTV at US$29.3B and social video at US$31.9B, per IAB. iab.com ↗
MarketSizeYearSource
Global gamesUS$201.6B2025Newzoo
Global games (forecast)US$213.9B2026Newzoo
Global box officeabout US$33.6B2025Gower Street
OTT streaming videoUS$169B2024PwC
Global advertisingUS$1.14T2025WPP Media
US digital video ads (forecast)US$81.9B2026IAB
Creator economyUS$250B2023Goldman Sachs
China micro-dramaRMB 67.79B2025CNSA

Nearly half of China mini-game ad creatives are now AI-made

As of 17 Oct 2026 · 6 sources

Mobile game user acquisition is one of the highest-volume buyers of video ads and among the fastest adopters of AI creative. Game publishers already commission drama-style ad series, and AI short-drama audiences increasingly resemble game UA audiences. Data providers caution that fully AI-generated creatives can raise ad costs.

  • In H1 2026, 46% of China mini-game ad creatives were AI-made across 26.8M+ creatives, 63% of them video, per DataEye. gamenewstc.com ↗
  • Gaming app UA spend was US$25B in 2025, and top gaming advertisers ship 2,400 to 2,600 creative variants per quarter, per AppsFlyer. appsflyer.com ↗
  • Video accounted for 53.7% of mobile game ad formats, per Sensor Tower. sensortower.com ↗
  • Sett, an AI agent platform for game video and playable ads, raised a US$30M Series B with customers including Zynga, Playtika and Papaya. siliconangle.com ↗
  • AI short-drama audiences skew 48% male, aged 18 to 34, iOS and Tier-1 cities, per 36Kr. 36kr.com ↗
  • Last War's Q4 2024 soap-opera style ad campaign drove more than 12.5M downloads, per Deconstructor of Fun. deconstructoroffun.com ↗

Platforms and distribution

AI-native drama apps enter the US top-grossing 30

As of 17 Oct 2026 · 4 sources

Overseas short-drama apps continue to grow in-app revenue, and AI-native apps now appear in US top-grossing charts alongside incumbents. Established platforms report a rising AI share of content. Unit economics remain difficult, with most projects reported as unprofitable.

  • Overseas short-drama apps earned about US$750M in in-app purchases in Q1 2026 (+20% YoY) with over 850M downloads (+140%); DramaBox and ReelShort took about US$140M each, per Sensor Tower. sensortower.com ↗
  • AI accounts for about 35% of ReelShort content and about 40% of DramaBox content; North American cost per install is US$5.28 and 80% to 90% of projects lose money, per The Paper. thepaper.cn ↗
  • Holywater's My Muse uses Google Veo 3 for AI series, per a Google case study. aistudio.google.com ↗
  • Character.ai launched an AI microdrama division, per The Hollywood Reporter. hollywoodreporter.com ↗

The AI microdrama tool category is already crowded and funded

As of 13 Sep 2026 · 6 sources

Multiple funded products in China and the West now market themselves explicitly as AI tools for microdrama, many claiming cross-episode character consistency. Generic microdrama generators are widely available, including open-source pipelines. Long-arc consistency across a full season is still reported to require manual policing.

  • StoReel launched Canvas, an AI production platform targeting microdrama creators, backed by a US$34M seed and user-acquisition round, per Variety. variety.com ↗
  • Morphic announced a Character Lineup workflow claiming to cut microdrama production busywork by up to 70% while ensuring series consistency. prnewswire.com ↗
  • Vigloo launched a beta of Vigloo Studio; its 22-episode Bloodbound Luna was produced AI-natively in 8 weeks by fewer than 10 people. globenewswire.com ↗
  • Character.AI entered microdrama in July 2026 with series whose characters can talk back, per Forbes. forbes.com ↗
ProductCompany / regionPositioningDisclosed funding
CanvasStoReel (West)Microdrama production plus D2C appUS$34M
MorphicMorphic (West)Series consistency for microdramaNot stated
Vigloo StudioVigloo (West)Studio workflow turned creator tool (beta)Not stated
Popshort.AIPopshort (West)One-person AI drama factoryNot disclosed
SkyReelsKunlun Wanwei (China)AI short-drama platform, open-source modelsNot stated
Dreamina / JimengByteDance (China)AI Microdrama Maker tied to Douyin and HongguoPart of ByteDance
Kling 3.0Kuaishou (China)Multi-shot consistency for short dramaPart of Kuaishou

Funding and valuations

Higgsfield and Runway lead AI video valuations above US$5B

As of 31 Aug 2026 · 1 sources

Valuations across the AI video stack span from multi-billion-dollar model makers and aggregators to seed-stage filmmaking tools. Several revenue figures are founder-stated or third-party estimates rather than audited disclosures. Aggregators and avatar platforms report revenue comparable to or larger than many model makers.

    CompanyLayerLatest round / valuationDateReported revenue metric
    HiggsfieldAggregatorUS$400M Series B at US$5.4BAug 2026about US$700M annualized (founder-stated)
    RunwayModelUS$315M Series E at US$5.3BFeb 2026Estimates conflict
    MiniMax (Hailuo)ModelHK IPO, US$619M raisedJan 2026US$79M (2025)
    Luma AIModelUS$900M Series C, about US$4BNov 2025Not stated
    SynthesiaAvatar platformSeries E at US$4BJan 2026over US$100M ARR
    HeyGenAvatar platformUS$500M valuationn/aabout US$200M ARR
    KreaCreative suiteUS$47M Series B, about US$500MApr 2025about US$8M ARR (Apr 2025)
    CaptionsEditingUS$75M growth round, about US$500MMar 2026Not stated
    Vidu (Shengshu)ModelUS$290M raiseApr 2025Not stated
    FlickFilmmaking toolUS$6M seedMay 2026Not disclosed
    SettGame ad agentUS$30M Series BMar 2026Not stated
    HolywaterDrama appUS$22M roundJan 2026Not stated

    Claim ledger

    81 claims from public articles, dated by the event, refreshed weekly from SOPHIA.

    MethodOnly claims at SOPHIA confidence 0.75 or above with a single clean statement are listed; 194 lower-confidence or unparsed extracts are held back, and 18 interview excerpts appear in their own section below (293 in total). Confidence is shown on each row.

    • 3 Sep 26

      Seedance

      ByteDance Seedance holds >80% market share by daily compute consumption in Chinese AI video market and 95% penetration in short-drama industry as of late 2026.

    • 1 Sep 26

      Veo

      Google Veo 3.1 (current version as of September 2026) API pricing: Veo 3.1 Lite at $0.03-$0.05 per output second (720p, best for drafts); Veo Fast at $0.10 per second (720p); Veo Standard at $0.40 per second (720p or 1080p). Free tier offers 50 credits/day in Google Flow AI Studio. Subscription access via Google One AI Premium at $19.99/month.

    • 1 Sep 26

      Veo

      Google Veo 3.1 generates videos up to 8 seconds in length at 720p/1080p with native audio synchronization, advanced physics simulation, and camera control parameters. Luma Agents integrates Veo 3.1 as a partner model.

    • 1 Sep 26

      Pika Labs

      Pika Labs has raised $135 million across 2 rounds (Series B $80M at $700M valuation, June 2024) and is estimated at ~$900 million valuation in early 2026, with projections above $1.5 billion by end-2026. No confirmed Series C round as of September 2026.

    • 1 Sep 26

      Pika

      Pika 2.2 (2026) introduced: improved motion consistency, video generation up to 10 seconds at 1080p resolution, 'Pikaframes' keyframe transition technology for smoother scene transitions, better prompt adherence, and improved lip-sync accuracy. Pika 2.5 adds cinematic effects, new storyboard mode, and additional PikaTwist effects.

    • 17 Aug 26

      Higgsfield

      Higgsfield raised $400 million in a Series B at a $5.4 billion valuation on August 17, 2026, led by DST Global with Goldman Sachs Growth Equity, Tribe Capital, Smash Capital, Fifth Wall, Valor Capital, Intel Capital, Mirae Asset, and NTT DOCOMO Ventures. Prior: $80M Series A extension led by Accel at $1.3B (Jan 2026); $50M Series A at $1.0B (Sep 2025).

    • 17 Aug 26

      Higgsfield

      Higgsfield's annualized revenue reached $700 million by August 2026, up from approximately $200 million at the end of 2025. The company has 30 million+ global users across 238 countries/territories, 390 Fortune 500 customers, and generates 20 million+ content pieces monthly.

    • 6 Aug 26

      ReelShort

      ReelShort's 2026 profitability was structurally enabled by its web-store bypass: routing premium users to its own web checkout to avoid Apple/Google commissions of up to 30%. This payment infrastructure — not content investment alone — is the primary identified driver of first meaningful net profit (~$40M) in 2026. Source: TechTimes, August 6, 2026.

    • 6 Aug 26

      ReelShort

      ReelShort is the first short drama platform to demonstrate the category can scale to $1B+ revenue, attract paying subscribers in volume, and generate positive net margins — provided the platform builds web-checkout payment infrastructure to bypass app-store commissions. Source: TechTimes, August 6, 2026.

    • 1 Aug 26

      ReelShort

      ReelShort (Crazy Maple Studio / COL Group) is on track for $1.05 billion gross consumer spend in 2026 — its first year past $1B — and first meaningful net profit of ~$40 million (EBITDA ~$63 million), after an estimated $12 million net loss in 2025. Revenue trajectory: $97M (2023) → $400M (2024) → $785M (2025) → $1.05B projected (2026). Source: Variety citing Media Partners Asia, August 2026.

    • 1 Aug 26

      Holywater

      Holywater CEO Bogdan Nesvit (August 2026): major studios and established creators are proactively approaching Holywater seeking content deals — a reversal from 2024 when Holywater had to seek them out. This signals mainstream Hollywood's recognition of microdrama as a serious distribution channel. Source: Variety, August 2026.

    • 1 Aug 26

      MiniMax

      MiniMax serves 300 million+ users across 200+ countries and 1 million+ enterprises across 100+ countries as of mid-2026. Sacra estimates $800M annualized revenue as of August 2026, up from $100M at end of 2025.

    • 31 Jul 26

      Seedance

      ByteDance released Seedance 2.5 on July 31, 2026, the first major AI video model to generate a native 30-second 4K clip in a single pass. Key new capabilities: up to 50 simultaneous multimodal reference inputs (up from 12 in Seedance 2.0), and local editing that allows modification of specific scene areas without regenerating the full 30-second clip.

    • 23 Jul 26

      FlovaAI

      FlovaAI raised $80M angel round (HSG, IDG, Yunji Capital) for a company founded less than one year prior. Founded by Guo Lie, previously co-incubated FaceU and Jianying (CapCut's predecessor) at ByteDance.

    • 13 Jul 26

      Aishi Technology

      Aishi Technology (maker of PixVerse) closed cumulative Series C+ funding of approximately CNY 2.98 billion (~$440 million USD) in July 2026, led by CDH Investments with Alibaba leading a follow-on extension. Valuation exceeded $2 billion. Initial Series C (March 2026) was approximately $300M. ARR exceeded $40 million by end of 2025.

    • 13 Jul 26

      PixVerse

      PixVerse had 150 million+ registered users and 15 million+ monthly active users as of July 2026. API pricing is $4.80 per minute for image-to-video generation. The platform offers three product lines: V-Series (consumer/API), C-Series (professional/commercial), R-Series (world models for game development).

    • 6 Jul 26

      Shengshu Technology

      Shengshu Technology raised $500M in Series B+ in July 2026, bringing total cumulative financing to over 5 billion RMB across 3 rounds in 5 months. Pre-IPO shareholding restructuring completed by end of March 2026.

    • 2 Jul 26

      Kling AI

      Kling AI raised $2.8 billion in Series A financing at an $18 billion post-money valuation, the largest single AI video funding round ever disclosed.

    • 2 Jul 26

      Kling AI

      Tencent, Alibaba Cloud, and Baidu all co-invested in Kling AI's $2.8B round simultaneously — the first time all three have backed the same AI deal.

    • 1 Jul 26

      Seedance

      ByteDance Seedance 2.0 (launched February 2026) holds >80% of China's AI video market share by daily token consumption volume and 95% penetration in the short drama industry, with monthly revenue exceeding 1 billion RMB.

    • 1 Jul 26

      Seedance

      ByteDance Seedance 2.0 operates at approximately 70% gross margin due to a mixture-of-experts (MoE) architecture.

    • 1 Jul 26

      My Muse

      Holywater's My Muse (AI streaming app) achieves near-identical paying subscriber conversion to My Drama (live-action): 23.6% on My Muse vs 23.9% on My Drama in May 2026 survey data. 89% of My Muse users want more AI-generated stories. This validates that AI-generated short drama content can match live-action monetization rates. Source: Holywater/Owl & Co 'State of Microdrama 2026' report, July 2026.

    • 1 Jul 26

      My Drama

      My Drama (Holywater): male share of monthly active users grew 28-fold over seven quarters, from 1.1% to 30.3%. My Drama was named 'Best Streaming Service' People's Voice at the 2025 Webby Awards. Source: Holywater/Owl & Co 'State of Microdrama 2026' report citing proprietary product data, July 2026.

    • 1 Jul 26

      DramaWave

      DramaWave (Kunlun) AI transformation: over 90% of new monthly content on the platform is AI-generated as of July 2026. DramaWave launched its AI comic-drama channel at the 30th Hong Kong International Film & TV Market in March 2026. This makes DramaWave the most AI-native of the major overseas short drama platforms. Source: BigGo Finance, August 2026.

    • 1 Jul 26

      Kling AI

      Kling AI (Kuaishou's video AI unit) raised nearly $3 billion in a funding round led by General Atlantic with Tencent participating, at an $18 billion post-money valuation, as of July 2026. This was a spinoff from Kuaishou with a planned independent HK Stock Exchange IPO filing in early 2027.

    • 1 Jul 26

      Kling AI

      Kling AI had over 60 million global users and more than 100 million cumulative users as of mid-2026. The company ranks 4th globally in text-to-video models.

    • 30 Jun 26

      Kunlun Tech

      Kunlun Wanwei (Kunlun Tech) committed $500 million to overseas short drama investment in December 2024. Kunlun Tech H1 2026 total revenue: RMB 5.36 billion (~$736M), with overseas revenue representing 94% of total. AI short-drama monthly gross revenue exceeded $65 million by H1 2026. Source: BigGo Finance citing Kunlun Wanwei financials, August 2026.

    • 30 Jun 26

      DramaWave

      DramaWave AI comic drama module (launched H2 2025): monthly active users reached tens of millions; monthly revenue in millions of dollars. Hit AI comic drama 'Seize the Dictionary! I Transferred to the Undead Calamity' had production cost <$20,000 and generated millions of dollars in revenue — demonstrating >100x ROI on AI-native content. Source: BigGo Finance, August 2026.

    • 30 Jun 26

      Kling AI

      Kling AI's Q2 2026 revenue was above RMB 850 million (approximately $125 million), up more than 200% year-over-year. Approximately 75% of Kling AI's revenue comes from overseas markets including North America.

    • 30 Jun 26

      MiniMax

      MiniMax's H1 2026 total revenue was US$116.6 million, up 283.1% year-over-year from US$30.4 million, exceeding the full-year 2025 revenue of US$79.0 million in just six months. Open platform and enterprise AI services revenue surged 703.1% to US$73.9 million (63.4% of total). AI-native products (including Hailuo AI) grew 100.9% to US$42.6 million.

    • 30 Jun 26

      MiniMax

      MiniMax H1 2026 gross profit was US$20.8 million (+464.8% YoY), with gross margin improving to 17.9% from 12.1% in H1 2025. Net loss for the period was US$357.997 million; adjusted net loss US$293.0 million. R&D expenses: US$296.9 million (+138.8%).

    • 1 Jun 26

      HeyGen Inc

      HeyGen reached $200 million in annual recurring revenue (ARR) as of June 2026, doubled from $100 million eight months earlier. The company remains cash-flow break-even having burned only $25 million of its total $74 million raised.

    • 1 Jun 26

      HeyGen Inc

      HeyGen's capital efficiency metrics as of June 2026: revenue per dollar raised = $2.70; revenue per employee = $1.5 million annually. Total funding: $74 million across 3 rounds (Series A $60M led by Benchmark at $500M valuation, June 2024).

    • 1 Jun 26

      HeyGen

      HeyGen has 30 million+ users who have created approximately 120 million videos across 250+ languages as of June 2026. 85% of the Fortune 100 are represented among users. Primary customer segments: small business owners, real estate agents, personal trainers, online course instructors.

    • 1 Jun 26

      HeyGen

      HeyGen's latest model Avatar V generates videos 10x more efficiently than Google Veo 3.1 while using only 1-2% of total video data compared to competitors. The company focuses on 'message-first video' with digital twin avatars rather than generalist video generation.

    • 1 Jun 26

      Moonvalley

      Reka AI acquired Moonvalley (maker of Marey AI video model) in an all-share exchange deal in June 2026. Moonvalley's co-founders Mateusz Malinowski and Mikołaj Bińkowski — both former Google DeepMind staff research scientists and key contributors to Veo — joined Reka's research leadership.

    • 1 Jun 26

      Adobe

      Adobe Firefly reached 6 million+ monthly active users by mid-2026, a 65% increase year-over-year. It is used by 75%+ of Fortune 500 companies. Enterprise contracts account for 61% of total Firefly revenue. The Firefly-powered asset marketplace generated $50 million in revenue from asset sales and licensing fees in 2026.

    • 1 May 26

      Higgsfield

      Higgsfield launched its flagship 'Supercomputer' product in May 2026, with agentic products growing 42-fold in the three months after launch. Key products: Ads 2.0, Soul UGC Builder, UGC Factory for e-commerce creative teams. Primary verticals: advertising, marketing, media/entertainment, fashion, retail, pharmaceuticals.

    • 30 Apr 26

      DramaWave

      Kunlun Tech's DramaWave platform achieved $128M monthly gross billings in April 2026, topping all A-share listed companies in overseas short-drama expansion, with ARR exceeding $570M.

    • 30 Apr 26

      ByteDance

      China's domestic AI short-drama market has a ~90% industry loss rate in H1 2026; revenue-sharing rates on ByteDance's Hongguo platform fell 90% in April 2026 (from 30-100 yuan to 5-10 yuan per 10K views).

    • 30 Apr 26

      DramaBox

      DramaBox (Dianzhong Technology / StoryMatrix) gross billings April 2026: $74 million (third among Chinese short-drama platforms going overseas). Revenue trajectory: $8M (2023) → $217M (2024) → $276M (2025). YoY growth 2023→2024: 2,550%. Source: Filmustage 2026 comparison citing Sensor Tower / industry data.

    • 30 Apr 26

      PineDrama

      PineDrama (ByteDance) is free-to-watch with no per-episode paywall, monetised through TikTok's advertising and revenue-sharing infrastructure. By April 2026, TikTok's monthly drama revenue-sharing payouts hit $15 million (60%+ of its entire Q1 total). Source: KR-Asia + Felix Braberg Substack analysis, 2026.

    • 30 Apr 26

      DramaWave

      DramaWave (Kunlun Wanwei) April 2026 monthly turnover: $128 million — officially #1 among A-share short drama platforms going overseas, surpassing Zhongwen Online. DramaWave total MAU ranked #1 in the overseas short drama market as of H1 2026. DramaWave/Kunlun displaced ReelShort and DramaBox from #1/#2 positions in H1 2026. Source: BigGo Finance citing Kunlun Wanwei financials, August 2026.

    • 30 Apr 26

      DramaWave

      DramaWave monthly turnover trajectory: end of March 2026: $48M monthly (ARR >$570M) → April 2026: $128M monthly — representing 2.67x growth in one month, driven by AI comic drama surge. Source: BigGo Finance citing Kunlun Wanwei financials, August 2026.

    • 26 Apr 26

      Kling AI

      OpenAI discontinued the Sora consumer app and web experience on April 26, 2026, after announcing shutdown on March 24. Sora's total lifetime revenue was approximately $2.1 million against ~$1 million/day operating cost.

    • 13 Apr 26

      Vidu

      Shengshu launched Vidu Q3 in April 2026, featuring reference-to-video capabilities with expanded visual and audio generation. The company's strategic direction is a general-purpose world model connecting digital game worlds with physical-world applications in autonomous driving and robotics.

    • 10 Apr 26

      Shengshu Technology

      Shengshu Technology (maker of Vidu) completed a Series B of approximately CNY 2 billion (~$293 million USD) led by Alibaba Cloud in April 2026, with Andon Haitang, China Internet Investment Fund, TAL Education Group, and Luminous Ventures participating. Series A+ (Feb 2026) had raised RMB 600M+. Total funding: $987M.

    • 10 Apr 26

      Shengshu Technology

      Shengshu Technology is exploring a Hong Kong IPO to raise more than $500 million, working with China International Capital Corp. (CICC) and Citic Securities. The company has fewer than 100 employees and was founded in March 2023 by Tsinghua University professor Zhu Jun.

    • 1 Apr 26

      FreeReels

      FreeReels (Skywork AI PTE.LTD., Singapore) surpassed 100 million cumulative downloads in Q1 2026 with a 137% download surge. By August 2026, FreeReels reached 230 million total downloads. FreeReels ranked #8 globally by app downloads in Q1 2026 — the highest-ranked short drama app globally that quarter. Source: Tiger Brokers citing app store data, Q1 2026.

    • 1 Apr 26

      Synthesia

      Synthesia crossed $100M ARR in April 2025 and is tracking toward $200M+ revenue for full-year 2026. Net revenue retention exceeded 140% as of April 2026. Contracts above $100,000 tripled over the prior 12 months.

    • 31 Mar 26

      Kling AI

      Kling AI's ARR grew from $100M in March 2025 to $500M in March 2026, a 4x year-over-year increase, with approximately 75% of revenue from overseas markets.

    • 31 Mar 26

      Kling AI

      Kling AI's Q1 2026 revenue exceeded 650 million RMB, growing more than 300% year-over-year.

    • 31 Mar 26

      Aishi Technology

      Aishi Technology (PixVerse) reached 150 million global users, 15 million MAU, and $40M ARR as of March 2026. Its model training cost is approximately 10% of industry peers.

    • 31 Mar 26

      Kling AI

      Kling AI's Q1 2026 revenue exceeded RMB 650 million (USD 95.8 million), up more than 300% year-over-year. The March 2026 annualized revenue run rate was approximately USD 500 million, 4x the level recorded a year earlier.

    • 25 Mar 26

      StoReel

      StoReel raised $34M ($9M seed led by Play Ventures + $25M user-acquisition financing from PVX Partners) and can produce an hourlong AI micro-drama series for $20,000-$40,000, versus $150,000-$200,000 for live-action.

    • 24 Mar 26

      Mirage

      Mirage (the company behind Captions AI video-editing app) raised $75 million in growth financing from General Catalyst in March 2026, bringing total funding to $175 million. The company has 20 million+ global users, 200 million+ videos created, $28.4 million in in-app revenue (trailing 12 months), and 3.2 million downloads in the prior 365 days.

    • 24 Mar 26

      Mirage

      Mirage rebranded from 'Captions' in 2026 to position itself as an AI lab serving advertising and marketing verticals. It switched to a freemium model in January 2025 to compete with ByteDance's CapCut and Meta's Edits app, and is expanding into Asian markets.

    • 19 Mar 26

      fal.ai

      fal.ai's annualized revenue doubled from approximately $200 million to $400 million between October 2025 and March 2026. As of March 2026, fal.ai was in discussions to raise $300M-$350M at an $8 billion valuation, nearly doubling from the $4.5B valuation achieved in December 2025.

    • 19 Mar 26

      fal.ai

      fal.ai serves 3 million+ developers and 300+ enterprise customers including Adobe, Canva, Amazon MGM Studios, and Shopify, processing 100 million+ inference requests daily across 28 global data centers. The platform provides serverless GPU inference for image, video, audio, and 3D generation models.

    • 1 Mar 26

      Luma Agents

      Luma launched Luma Agents in March 2026, a multimodal agentic creative platform that wraps Luma's own models with partner models including Google Veo 3.1, Kling, and Seedance to handle end-to-end creative projects across text, image, video, and audio.

    • 1 Mar 26

      Adobe Firefly Video

      As of March 2026, Adobe integrated 12 third-party generative video models within the Firefly interface. Adobe Firefly Video generates cinematic b-roll clips up to 5 seconds, trained on licensed content for commercial safety. Runway partnership was announced in December 2025.

    • 12 Feb 26

      Seedance

      ByteDance launched Seedance 2.0 on February 12, 2026 — a multimodal AI video generation model that accepts text prompts, reference images, and audio clips and produces HD video in a single unified generation pass.

    • 10 Feb 26

      Runway

      Runway raised $315 million in Series E financing in February 2026 and expanded into world models, robotics, and gaming. Its Gen-4.5 model leads independent text-to-video benchmarks as of mid-2026.

    • 10 Feb 26

      Runway

      Runway raised $315 million in a Series E round at a $5.3 billion valuation in February 2026, led by General Atlantic with investors including Nvidia, Fidelity Management & Research, AllianceBernstein, Adobe Ventures, Mirae Asset, Emphatic Capital, Felicis, Premji Invest, and AMD Ventures. Previous round: $308M Series D at $3.3B (April 2025).

    • 10 Feb 26

      Runway

      Runway's enterprise customer base grew 300% in the past 12 months as of early 2026, with named customers including Yamaha, SoftBank, NHN (Japan), BBC, Fremantle, and WPP (Europe). Developer platform customers include Adobe, ElevenLabs, Shutterstock, Figma Weave, and Gamma.

    • 1 Feb 26

      Kling AI

      Kling 3.0 (launched February 2026) generates native 4K video at 60fps using a DiT architecture with 3D VAE — no upscaling — with clips up to 15 seconds.

    • 1 Feb 26

      Kling

      Kling 3.0 launched in February 2026 with multimodal capabilities. Primary professional use cases are advertising/marketing, film/TV/short drama, and gaming. The $3B raise is earmarked for data center expansion and talent recruitment.

    • 31 Jan 26

      PineDrama

      ByteDance's multi-layered global short drama expansion strategy: (1) Melolo — SEA (Indonesia/Philippines), free/IAA model, launched Nov 2024; (2) PikoShort — Japan market; (3) PineDrama — US and Latin America, free/ad-supported, launched January 2026, 17.6–17.8M downloads in first 30 days; (4) TikTok Minis — embedded within TikTok. Source: KR-Asia citing Vitrina.ai, 2025–2026.

    • 26 Jan 26

      Synthesia

      Synthesia raised $200 million in a Series E at a $4 billion post-money valuation in January 2026, led by GV (Google Ventures) with NVentures (Nvidia), Kleiner Perkins, Accel, NEA, Evantic, and Hedosophia. Previous valuation was $2.1B one year prior. Total funding: $536.6M across 7 rounds.

    • 26 Jan 26

      Synthesia

      Synthesia enterprise penetration as of early 2026: 90% of Fortune 100 companies and 95% of the DAX 40 are customers; 65,000+ total global customers; 70% of revenue from enterprise deals. Named customers include Bosch, Merck, and SAP.

    • 26 Jan 26

      Synthesia

      Synthesia is expanding from AI video creation into AI agents for employee training and knowledge management, positioning agents as its 'core strategic focus' in 2026. The company has 500+ employees across London HQ and offices in Amsterdam, Copenhagen, Munich, New York, and Zurich.

    • 15 Jan 26

      Holywater

      Holywater raised a $22 million Series A in January 2026, led by Horizon Capital with strategic participation from Endeavor Catalyst and Wheelhouse (Brent Montgomery's production company) — described as a 'record' raise for a microdrama company at that stage. Fox Entertainment previously took an equity stake (undisclosed terms). Source: Axios, January 15, 2026.

    • 15 Jan 26

      Holywater

      Holywater's content ecosystem (My Drama + My Passion + My Muse) reaches over 100 million users. Revenue grew approximately 3x in 2025. Estimated annual revenue ~$70 million across the portfolio as of January 2026. Source: Holywater official press release, January 2026.

    • 13 Jan 26

      PixVerse

      Aishi Technology launched PixVerse R1 in January 2026, described as the world's first general-purpose real-time world model supporting up to 1080P resolution. R1 is an Omni multimodal model that integrates text, images, audio, and video in one unified processing pipeline — unlike separate-pipeline competitors. Supports up to 4K video generation with embedded audio.

    • 1 Jan 26

      DramaBox

      DramaBox won Google Play's Best Entertainment App Award for 2024, and its net profit margin for its short-drama business is less than 1% — despite being the only confirmed profitable major short drama platform at scale. Source: Filmustage 2026.

    • 1 Jan 26

      Holywater

      Holywater (My Drama + My Passion + My Muse) reached approximately $70 million in annual revenue as of early 2026 — a $0-to-$70M revenue build using mobile subscription infrastructure via RevenueCat. Revenue grew ~3x in 2025 alone. Source: RevenueCat customer case study, 2026.

    • 1 Jan 26

      ShortMax

      ShortMax achieved $17M+ monthly revenue with 30M+ peak MAU. Top geographic markets: Indonesia (16.62%), Thailand (9.82%), Philippines (8.19%) of downloads. May 2025: 279% month-on-month revenue surge. ShortMax is also a content supplier to ByteDance's PineDrama, giving its catalog a second distribution window inside the TikTok ecosystem. Source: Sensor Tower 2026 report.

    • 1 Jan 26

      NetShort

      NetShort holds 10% of the ex-China short drama market (per MPA 2026). NetShort charges approximately $29.99/week, generating estimated ~$180M ARR. NetShort grew 10% YoY and is emerging as a strong IAP challenger in Japan, South Korea, and Southeast Asia. Source: Sensor Tower 2026 + Streaming Lens data.

    • 1 Jan 26

      ShortMax

      ShortMax catalog: 50,000+ short dramas and movies in 19 languages across 200+ countries. Production philosophy: mass-produce at $12K–$50K per Chinese-origin series (7-day shoot, 100 episodes), localize/dub, release daily — a portfolio volume strategy. Source: AXIS AI Studios 2026 citing industry data.

    • 1 Jan 26

      Dream Machine

      Luma AI's Ray 3.14 model (released January 2026) generates video at native 1080p output at 4x the speed of earlier versions, produces clips up to 18 seconds at 24 fps, and supports native 16-bit HDR with ACES EXR output.

    • 1 Jan 26

      MiniMax

      MiniMax completed a Hong Kong IPO in January 2026, raising approximately $619 million at a $4 billion valuation. Investors include Alibaba and Tencent. The company's cash balance was $1,322.8 million as of June 2026.

    Interview excerpts

    Statements from recorded interviews and podcasts in SOPHIA, shown with the original video.

    MethodExcerpts are machine transcribed and kept as spoken, tagged with the company they concern. They may include the interviewer's lines and are statements, not verified facts.

    Recorded interview

    YouTube · 4 Aug 2026 · 1 excerpt

    • TikTokAnd when people are really bought into the story, and they're they're usually engineered in in a really smart, deliberate where it's, you know, something's building up, someone is getting wronged, and finally, they're about to reveal to the world that they're someone awesome, and then you have to pay. [laughter] Discovery-wise, I mean, cuz there are the paid micro-drama platforms, but is most of the discovery happening on traditional short platforms like TikTok, Instagram Reels, and then they kind of go into one of the paid platforms? Yeah, largely, Facebook is really big, TikTok is really big. Okay, and then it pushes them into one of the other. That's right.

    Recorded interview

    YouTube · 25 Jun 2026 · 1 excerpt

    • SynthesiaBefore hooking up with 90% of the Fortune 100 Synthesia got its start making videos of David Beckham speaking 9 languages, Lionel Messi speaking 8 languages, and Snoopy Dogg rapping in Australian.

    Inspired Capital

    YouTube · 6 May 2026 · 6 excerpts

    • RunwayWell, Cristóbal Valenzuela is the co-founder and CEO of Runway, the company that pioneered AI video generation and is building world models that will power the next era of human storytelling.
    • RunwayAnd in so many ways you actually um you built Runway well before generative AI took off on the main stage.
    • RunwaySo, language models and coding models and all of them have benefited from those three variables kind of becoming better over time. In the last few years, what has surprised you about the progress of Runway?
    • RunwayUm first, tell us what is the business model of Runway and how did you prioritize building this truly groundbreaking product, which is anybody can create incredibly high-quality generative uh video.

    Semafor

    YouTube · 1 May 2026 · 1 excerpt

    • RunwayUh and maybe for context, I've started working on Runway quite some time ago before AI was actually like working in the way it's working right now.

    Sequoia Capital / Training Data

    YouTube · 1 May 2026 · 1 excerpt

    • BenchmarkBut that's where the interesting combination will happen between like what's the quality versus latency uh benchmark that um that you that you have.

    The Upstarts Podcast (Alex Konrad)

    YouTube · 1 May 2026 · 1 excerpt

    • RunwayRunway is an applied research company that builds video and world models.

    TechCrunch Equity

    YouTube · 10 Apr 2026 · 4 excerpts

    • Luma AIBig names like Yann LeCun and Fei-Fei Li are betting on world models, but Luma AI founder Emmett James says that's still small thinking.
    • RunwaySo the world models being developed by World Labs, being developed by Runway, and these kind of companies, they're like, you know, just very lazy attempts at trying to make a video model interactive.
    • TikTokWhat Netflix actually proved, it's a Netflix, TikTok, and now Instagram Reels are in the same bucket because of that.
    • TikTokSays actually TikTok and Instagram's growth.

    Recorded interview

    YouTube · 3 Mar 2026 · 2 excerpts

    • HiggsfieldA lot of businesses can really scale to tens of millions of dollars today profitably with AI. This is Alex, founder of Hexafield, an AI company that hit a $200 million annual recurring revenue in just 9 months, faster than Slack or Zoom. And what he told me about starting a business today completely blew my mind, and you can copy his strategy, too. Focus on bringing like the first dollar by day 30 of product development, and maybe 1 million ARR by day 90. That's a lot. I just think it's the next industrial revolution. It's probably more powerful than the internet. For many, many young people, AI becomes social elevator. For someone who still has fear, like this is moving so fast, I don't know how I can start. Can you give them one piece of advice? First, I would start from I have an amazing guest today. I am so excited to learn from you. Let's get very practical right away. You built Hexafield and achieved $200 million in revenue in 9 months. Let's imagine every I don't want this to happen, but let's imagine a scenario when you have to start from scratch tomorrow, and you have 90 days to launch a business idea. From what you've learned with your experience at Hexafie…
    • HiggsfieldSo, I recently spoke to a very large, very, very large like property management company, which actually uses Higgsfield to to sort of advertise their buildings, their uh apartments, and so on.

    Recorded interview

    YouTube · 1 Feb 2026 · 1 excerpt

    • TikTokNOW MORE COMPETITORS ARE ON THE NOW MORE COMPETITORS ARE ON THE SCENE, INCLUDING ONE RECENTLY SCENE, INCLUDING ONE RECENTLY LAUNCHED BY TIKTOK CALLED PINE LAUNCHED BY TIKTOK CALLED PINE

    Tool landscape

    59 AI video tools: what each does, its pricing and its reported funding and traction, with the date each row was checked.

    MethodEach row is compiled from the company's own site and pricing page plus press coverage, and dated when checked; the company name links to its site. LimitationPricing changes often and funding and traction figures are frequently company-reported; a dash means the figure was not found.

    CompanyProductWhat it doesPricingFunding and tractionAs of
    AbsurdAbsurdYC F25Online whiteboard/production canvas for AI film teams; also runs as an AI ad studio.Usage/spend based; plans not captured~$500K (YC, .406, Audacious) per third parties2026-10-14
    AdobeAdobe Firefly VideoGenerates 1080p video from text or still images with camera controls; Generative Extend in Premiere Pro; browser Firefly Video Editor (assemble AI + uploaded footage); AI audio (Generate Soundtrack/Speech); Prompt-to-Edit. Built for pro Creative Cloud workflows.Standalone Firefly: Free (25 cr), Standard $9.99/mo (2,000 cr), Pro ~$19.99–29.99/mo (4,000 cr, ~70 5s clips), Pro Plus $49.99/mo (10,000 cr), Premium $199.99/mo (50,000 cr, ~500 clips). Video ~20 cr/sec. CC subs get Firefly bundled; enterprise adds IP indemnity; API usage-based.N/A — public company (Adobe market cap ~$150–200B). No external rounds. Firefly direct subscription revenue est >$400M FY24–25; ~11% of new CC ARR FY2024; Adobe AI-influenced ARR >$5B FY2025. 24B total Firefly assets generated by Jun 2025 (mostly images; video launched Oct'24), ~1B/month. Paid Firefly subs ~doubled QoQ (Q4 FY2025). 72% of Fortune 500 design teams use Firefly. Enterprise: PepsiCo/Gatorade, IBM, Mattel, Deloitte, Paramount+. 32.5M CC subscribers reachable.2026-08-30
    AIspherePixVerseConsumer-first AI video generation: text/image/clip in → cinematic video out. Web + mobile (PixVerse globally; 'Paiwo AI'/拍我AI in China) + developer API. 13+ languages, 177 countries.Free: 90 signup + 60 daily renewable credits (720p, watermark). Standard ~$8/mo (1,200 cr), Pro ~$24/mo (6,000 cr, 1080p), Premium ~$48/mo (15,000 cr), Ultra ~$149/mo (25,000 cr). API from $100/mo. ~$4.80–5.40/min mid-tier.~$550M+ total. Series A ~$55M+ (Ant Group, Lighthouse, Beijing AI Fund). Series B $60M (Sep 2025, Alibaba lead). Series C $300M initial (Mar 2026, CDH lead — unicorn >$1B) → extended to $439M cumulative (Jul 2026, Alibaba + Mirae + BlueFocus). Post-money >$2B. ARR $40M+ (end 2025, 10x+ YoY). 150M+ registered, 15–16M MAU, 177 countries (Jul 2026). 2.1B+ videos generated. #1 photo/video app in 30+ markets; 6.2M downloads Jan 2025. 'Venom Effect' template 10B+ views (Nov 2024). TikTok ~652.9K / IG @pixverse_official ~87K / X @PixVerse_ ~20K / Discord ~99K.2026-08-30
    BlumaBlumaYC F25Clones winning UGC ads: de-edits a reference video and recreates every shot with AI.—YC standard; other rounds not disclosed2026-10-14
    ByteDanceDreamina / JimengOne-stop AI creative studio: text-to-image (Seedream), text/image-to-video (Seedance), digital humans, audio/music, action imitation. Web + mobile. Plugs into the ByteDance flywheel: Dreamina generates → CapCut edits → TikTok/Douyin distributes.Intl (Dreamina, USD approx): Free (~60–120 refreshing cr/day); Basic ~$15, Standard ~$35, Advanced ~$70/mo (annual -20–30%); $0.097/sec API. Domestic (Jimeng, CNY): Free (60–100 cr/day); Basic ¥79, Standard ¥239, Premium ¥649/mo. 3 domestic price hikes Mar–Apr 2026. Approx/dynamic. [reported]N/A standalone — a ByteDance product line. ByteDance valued ~$550B (Feb 2026); 2025 revenue ~$186B (overtook Meta quarterly Q2'25); historical raise $9.4B (Sequoia China, KKR, SoftBank), now self-funded; 2026 capex ~$70B (AI). [reported] Jimeng (China): WAU ~760K (Dec'24) → ~2M (Feb'25); ~6M registered by Apr'25 (65M+ videos, 175M+ images); MAU ~10.1M (Sep'25) → ~13.5M (Q1'26); #2 Chinese AI app by downloads Sep'25. Seedance 2.0 → 229% MoM download surge. Intl: 16M+ monthly site visits; inside CapCut's 170M–320M MAU. 'Jimeng Studio' film/TV label Aug 2026.2026-08-30
    ByteDanceSeedanceFoundation video-generation model family (text/image-to-video), up to 1080p@24fps, with native multi-shot storytelling (multiple camera cuts in one generation), strong motion stability, and multi-style range (realism/anime/film/ad). It is the ENGINE, not the app — image sibling is Seedream; consumer access via Jimeng/Dreamina/CapCut/Doubao.Volcano Engine API: 1.0 launch ~¥3.67 (~$0.50) per 5s 1080p; 1.5 Pro ~72% cheaper than 1.0 Pro; 2.0 token-based ~¥46/M tokens (≈¥1/second); 2.5 ¥70/M (no input) / ¥42/M (with video input). Consumer: free/subscription via Jimeng, Dreamina, CapCut, Doubao.Not independently funded — internal ByteDance/Seed product monetized via Volcano Engine's MaaS P&L. ByteDance is private (secondary valuations ~$300B+, context only). [structural fact; valuation UNCONFIRMED] #1 at launch on Artificial Analysis (Jun 2025), #2 currently (Aug 2026). Seedance 2.0 reportedly >RMB 1B/month revenue and >50% of Volcano Engine MaaS revenue; 2026 MaaS target ~RMB 15B (~$2.2B, ~10x 2025); 70–90% gross margin. Distributed via consumer apps + enterprise API. [rank; revenue/margin/ESTIMATE — 36Kr/kr-asia/BigGo]2026-08-30
    CardboardCardboardYC W26Browser-based agentic video editor: footage plus a goal gives a first cut in minutes.Lite $10 ($8 annual); Starter $40 ($29 annual); higher tiers exist; launch discountsYC standard; other rounds not disclosed2026-10-14
    Comfy OrgComfyUI / Comfy CloudOpen-source (GPL-3) node canvas for image and video models, monetised through Comfy Cloud, a credit wallet, Partner Nodes, Comfy Router, Comfy Agent and Comfy API.Cloud plans US$20 to US$700 a month; 211 credits = US$1; Comfy API GPU at US$3.49/hr.US$30M at US$500M valuation led by Craft Ventures (24 Apr 2026); about US$47M raised in total. Claimed 4M+ users, 60k community nodes, 150k daily downloads.2026-11-03
    DomoAIDomo AIVideo-to-video AI stylization: upload a clip → get anime/cartoon/watercolor/70+ styles with motion preserved. Also text-to-video, image-to-video, character animation, talking avatars (TTS + lip-sync up to 60s), face swap, 4K upscaling, background removal. Web + Discord.Free trial (15 cr). Basic $9.99/mo (500 cr, no watermark). Standard $27.99/mo (1,500 cr + unlimited Relax queue). Pro $69.99/mo (4,000 cr, 20–30s outputs, advanced lip-sync). Annual ~20–29% off. Non-expiring credit packs. No public enterprise tier.No public funding rounds (not in Crunchbase/PitchBook/Tracxn). Paid-up capital SGD 10,000. Shareholders 3→7 (Jul–Oct 2025; possible seed/angel, unconfirmed). Likely bootstrapped/revenue-funded. Valuation & revenue not disclosed. [NOT DISCLOSED] 5M+ active creators self-reported (Aug 2026; 4M+ in May 2026 PR). Discord ~1.35M members (#7 globally). Web ~118K visits/mo (Similarweb May'26 — Discord/mobile is primary surface). X @DomoAI_ ~18.9K. a16z Top 100 Gen AI Consumer Apps; covered by Japan's TV Asahi; Singapore SME 500 (2025).2026-08-30
    ElevenLabsElevenLabsAI voice/audio platform — TTS w/ emotion, cloning, dubbing 30+ langs, Scribe STT, conversational voice agents, Eleven Music, ElevenCreative (video)Usage-based + subscription + enterprise contracts; biggest single contract ~$2M (call center/support); tier prices not published$11B (Series D, $500M, Feb 2026); ladder $1.1B→$3.3B→$6.6B→$11B $600M total ARR (Jul 2026); $500M crossed May 2026; ~$330M end-2025; $200M Sep 2025. (Note: $193M is TTM recognized revenue Dec 2025, ~60% margin, profitable since inception —/Forbes)2026-08-20
    falfal.aiGenerative-media inference cloud: one REST/WebSocket API to run 600+ image/video/audio/3D models without managing GPUs. Billing per output or GPU-second; no subscription. Serves developers + enterprise (Adobe, Canva, Shopify, Amazon MGM).Pay-per-use, no subscription. Images ~$0.02–0.05 (Seedream V4 $0.03, FLUX Kontext Pro $0.04); video ~$0.05–0.40/sec. Custom GPU: H100 ~$1.89/hr, A100 ~$0.99/hr (per-second). Enterprise negotiated.~$587M total: seed $9M (a16z, 2022) → A $14M (Kindred, 2023) → B $49M (Feb 2025) → C $125M @ $1.5B (Meritech, Jul 2025) → $250M @ $4B+ (Kleiner+Sequoia, Oct 2025) → D $140M @ $4.5B (Sequoia, Dec 2025); reportedly in talks for $300–350M @ ~$8B (Q1 2026). Investors incl. NVIDIA, Bessemer, Shopify, Salesforce. ~$400M annualized revenue (Feb 2026, up from $25M end-2024; ~1,040% YoY per Sacra). 2.5M+ developers (May 2026); 50M+ creations/day. Enterprise: Adobe, Canva, Shopify, Perplexity, Amazon MGM. AWS preferred cloud provider (May 2026). Acquired Remade (Nov 2025). X @FAL ~32.8K.2026-08-30
    FamiliarFamiliarYC S26Real-time video dubbing that keeps the same actor's voice and lip movements across 25-30 languages.Free and paid tiers; exact prices not captured—2026-10-14
    FlickFlickYC F25'Figma + Cursor for AI filmmaking': idea to script to storyboard to shots to edit in one workspace, plus Flick TV for distribution.—$6M seed (May 2026): True Ventures, GV, Lightspeed, YC2026-10-14
    GenFlickGenFlickProsumer director agent that turns a story into consistent-cast shots across rented models (Seedance, Kling, Veo, Grok, Runway); publishes GenBench comparisons.Free tier; Starter US$19.90/mo (160 credits), Creator US$49.90, Pro US$200, Studio US$1,000.No funding found; appears bootstrapped (Oct 2026). Not disclosed; v1.65.0 shipped 10 Oct 2026.2026-10-14
    GenmoGenmoText-to-video generation: prompt → short (up to 5.4s, 30fps) photorealistic clip. Hosted playground (free + paid) or self-host via open weights. Optimized for cinematic/photoreal; weaker on animation. Layer = Brain + Hands (research lab + hosted product, Stability-AI style).Free: 250 lifetime credits (card required, not charged). Lite ~$10/mo (1,200 cr/mo, no watermark, commercial). Standard ~$30/mo (5,000 cr, priority, early access). 100 cr/Mochi video. Open weights free to self-host.$28.4–30M Series A (Oct 2024, NEA lead; The House Fund, Gold House, WndrCo, Eastlink, Essence VC; angels Amjad Masad, Abhay Parasnis). Valuation & revenue not disclosed. 1.5M+ users (~2024 interview); 80K daily at first product launch (Jan 2023). GitHub ~3,700 stars; HuggingFace ~3,368 downloads/mo, 1,350 likes. X @genmoai 14,983 / @ajayj_ 7,890. Captions app built on Mochi 1.2026-08-30
    GenrokGenrokURL-in, campaign-out AI video ad generator producing many variants for testing in 87 languages.Credits: Starter US$29 (240 cr), Creator US$79, Pro US$199, Growth US$499, Scale US$999; failed renders not charged.None found. Public launch 4 Nov 2026; users not disclosed.2026-11-05
    Google DeepMindVeo (Google)Text-to-video and image-to-video model family. Short cinematic clips (default 8s in-product) from text/image prompts, with camera control, real-world physics, and — as of Veo 3 — synchronized native audio. Delivered via the Gemini app, Flow filmmaking tool, YouTube Shorts, and Vertex AI / Gemini API.Consumer: Google AI Pro $19.99/mo (~1,000 cr, ~$0.25/s effective), AI Ultra $249.99/mo (~2,500 Veo 3.1 Fast videos); Flow included. Enterprise (Vertex AI, per-sec): Veo 3.0 $0.50/s video-only, $0.75/s with audio (≈$6 per 8s clip); Veo 3.1 $0.03/s (Lite 720p) to $0.50/s.None standalone — Veo is a product line inside Alphabet (NASDAQ:GOOGL), funded from Alphabet's balance sheet (~$350B 2024 revenue, multi-trillion mkt cap). No separate valuation/round/cap table. Videos generated in Flow: >100M by Aug 2025, >275M by Jan 2026 (Google-reported). Integrated into YouTube Shorts creator tools (Veo 3.1 'Ingredients to Video', Jan 2026). 2026 blind-preference arenas: Veo 3/3.1 win prompt fidelity + native dialogue + 4K short shots; Sora consumer app shut Apr 2026.2026-08-30
    GrocaloGrocaloYC S26AI brain that runs social content for big creators using video understanding and analytics.Not public (managed)YC standard; other rounds not disclosed2026-10-14
    HeyGenHeyGenAI avatar/spokesperson video generation + multilingual video translation/dubbing. Avatar videos from a script (stock or custom digital twin), translate video into 175+ languages with lip-synced voice cloning, interactive real-time avatars for sales/support/training. Frame: 'Create, Localize, Personalize.'Free (3 videos/mo, 720p) → Creator $29/mo (600 cr, 1080p, 175+ langs) → Pro $49–$4,300/mo (4K) → Business $149/mo +$20/seat → Enterprise custom. API pay-as-you-go from $5 min. Credits: Avatar III 3 cr/min; Avatar IV/V 20 cr/min.$74M total across 3 rounds: seed (HongShan + ZhenFund); $5.6M Conviction (Sarah Guo, Nov 2023); $60M Series A led by Benchmark (Victor Lazarte, board), + Thrive, BOND, SV Angel (2024) at $500M valuation. Angels: Dylan Field, Neil Mehta. Only $25M burned; cash-flow break-even 2026. $200M ARR (Jun 2026, doubled in 8 mo from $100M; profitable since Q2 2023). 30M+ users, 196 countries, 118M+ videos, 85% of Fortune 100. $1.5M revenue/FTE (~200 staff). Clients: McDonald's, Salesforce, Ryanair, Harvard HBS. vs Synthesia $146M ARR (Sept 2025) — HeyGen wins on growth, Synthesia on ACV.2026-08-30
    HiggsfieldHiggsfieldConsumer and prosumer AI video and image app that routes leading third-party models plus its own Soul image model, Cinema Studio and apps.Subscriptions plus credits; per-second API rates with one dollar balance (Sep 2026).Reported US$1.3B valuation (Jan 2026); reports of an US$8B valuation for a next round (Sep 2026). Claimed US$1B annualised run rate on 29 Sep 2026 (last 4 weeks x 13), 32M+ users, nearly 1M paying subscribers.2026-10-02
    InstaAgentInstaAgentYC P26 (Spring 2026)Scales one campaign into hundreds of persona-specific creatives and accounts on Meta and TikTok.Not public (agency/demo)YC standard; other rounds not disclosed2026-10-14
    InVideoinvideo AI / Agent TwoPrompt-to-video agent that writes, generates and edits end to end across 200+ models; Agent Two (Jul to Aug 2026) adds project memory, playbooks and workflows.Annual, per seat a month: Starter US$10 (400 credits), Plus US$30 (2,000 credits), Max US$75 (5,000 credits); Enterprise custom.50M+ users claimed; best public revenue figure about US$70M (Forbes India, Jun 2025).2026-10-16
    KaiberKaiberSuperstudio: infinite-canvas AI platform to generate/edit images + video from text, reference images, or uploaded audio. Audio-reactivity (animations sync to music) is a native primitive → dominant in AI music videos. Bundles Kling, Luma, Veo, Runway, Flux under one subscription.Flex (pay-as-you-go, free sign-up); Creator $29/mo (1,400 cr); Pro $149/mo (7,500 cr); Custom enterprise. Annual ~20% off; credit packs $5–$250. No perpetual free tier.Bootstrapped Nov'22–Oct'24 to 'seven figures gross revenue' + 5M+ signups before capital. Seed announced 16 Oct 2024, led by EQT Ventures w/ Crush Ventures; amount undisclosed (3rd-party ~$10M ESTIMATE). Valuation n/a. 5M+ creator signups ( Oct 2024). ~1/3 of DAUs paid at 2.5M-user mark (May 2023). ~$2.6M ARR est / 24 employees (Getlatka est, not founder-stated). 200K+ app downloads, 4.8 rating. X @KaiberAI ~22K; LinkedIn ~3,962; Discord 95K+ (2023, stale). Artists: Kid Cudi, Linkin Park, Grimes, Wu-Tang, Chief Keef.2026-08-30
    Kailash LabsKailash LabsYC F26Factory for small, specialised video reasoning models (distil frontier VLMs into cheap task models).Not publicYC standard; other rounds not disclosed2026-10-14
    Keyframe LabsKeyframe LabsYC P26 (Spring 2026)Real-time photoreal interactive avatars ('human foundation models') via embed code.Free 30 min; Starter $50/mo; Growth $100/mo; Scale $300/mo; extra from $0.06/minYC standard; other rounds not disclosed2026-10-14
    KoyalKoyalYC F25Turns a script or voiceover into cinematic scenes with consistent characters and settings.—~$500K-1M seed (Pioneer Fund, YC) per third parties2026-10-14
    KreaKreaCreative-AI suite — real-time, visually-controllable generation/edit/enhance of images, video, 3D; taste-and-control layer over every frontier model + own Krea 1/2Freemium tiers: Free $0 (100 units/day), Basic $9, Pro $35, Max $105, Business $200, Enterprise custom . Open-weights: only firms >$1M revenue pay a license~$500M post-money (Series B $47M, Bain, Apr 2025); $83M total raised $8M ARR (Apr 2025; "20x in 14 months"). 2026 ARR not disclosed2026-08-20
    KuaishouKlingAI video generation (text/image-to-video) via web + enterprise API. Clips up to 3 min, first/last-frame conditioning, cinematic camera controls, native audio (2.6), 4K/60fps (3.0), AI digital humans / virtual try-on for e-commerce.Intl (USD): Free (66 cr/day) → Standard $10 → Pro $37 → Premier $92 → Ultra $180/mo. API ~$0.42–4.20/generation. China RMB (灵感值) tiers align at ~7.2 RMB/USD. Credits expire monthly.Parent Kuaishou is public (HKEX:1024) — no independent Kling round as of Aug 2026. Kuaishou 2025 revenue RMB142.8B (+12.5%). Spinoff ~$18B pre-money target; a circulating $200B figure appears erroneous — discard. Kling FY2025 revenue RMB1.04B (~$150M); H1 2026 cumulative >RMB1.5B (~$208M), Q2 +200%+ YoY. ARR $240M (end25)→$300M (Jan26)→$500M (Apr26). Users 60M+ (end25)→100M+ (Jun26), 224 countries. Enterprise/API clients 30K→~50K. #1 design app in 40+ countries.2026-08-30
    Lamina Labs (Simi)Lamina Labs (Simi)YC P26 (Spring 2026)Near-real-time structured video; Simi turns prompts/docs into whiteboard explainer videos in seconds.Plus $19.99/mo; Pro $49.99/mo; Max $99.99/mo; Business customYC standard; other rounds not disclosed2026-10-14
    Leonardo.AiLeonardo.AiGenerative AI platform for image + video: generate/fine-tune/iterate via proprietary models (Phoenix, Lucid Origin) + integrated third-party (Veo 3, Kling 3.0, Sora 2). Tools: Realtime Canvas (sketch-to-image), Motion video, custom model training, Photoshop-style editor, enterprise team product.Solo: Free (150 tokens/day) / Essential $12 / Premium $30 / Ultimate $60 per mo (annual ~20% off). Team from $72/mo (3 seats). API pay-as-you-go from $5, no expiry. 'Unlimited' only on own models; premium 3rd-party (Veo/Sora/Kling) consume tokens.Raised ~$47M AUD (~$31M USD) pre-acquisition: seed ~$7.5M + Series A ~$23.5M (Oct 2023, lead Blackbird + Side Stage; Sierra, Smash, TIRTA, Gaorong, Samsung Next). Acquired by Canva 30 Jul 2024 for >=US$320M (Blackbird docs) / $349M (Dealroom) — officially undisclosed. 29M+ registered users end-2024, 30M+ by early 2026; 2B+ images generated. Web ~17.5M visits/mo (Semrush Apr'26). Discord ~1.58M members. X @LeonardoAi ~61K (unconfirmed). Revenue undisclosed (one 3rd-party est ~$95M 2026 ARR — non-audited, rough only). 120 employees at acquisition.2026-08-30
    LightricksLTX Studio / LightricksLTX Studio = end-to-end AI film-production workspace — turns scripts/prompts into characters, scenes, storyboards, and full sequences with directing controls (framing, camera, shot composition, timeline). Model-agnostic app routing to own LTXV/LTX-2 + third-party (Veo 3.1, FLUX.2, Kling, Seedance). Launched Feb 28 2024.Free ($0, 800 one-time cr); Lite $15/mo ($12 yr, 8K cr); Standard $35/mo ($28 yr, 28K cr, commercial license); Pro $125/mo ($100 yr, 110K cr, Veo 3.1); Enterprise custom. Credit-metered.~$335M raised Series A–D (A $10M 2015; B $60M 2018; C $135M 2019/unicorn; D $130M Sept 2021 at $1.8B valuation). $1.8B (2021) is the last public mark. Company-wide $250M ARR by Jul 2025 (mostly the mobile division); ~50M+ monthly users, ~6.6M+ paying, 730M+ total app downloads. LTX Studio-specific traction is not broken out.2026-08-30
    Luma AILuma (Dream Machine)Dream Machine — a multimodal AI creative workspace generating & editing video, images, audio in one agent-driven environment with shared context (no tool-switching). Consumer + enterprise tiers; API on AWS Bedrock + direct.Consumer: Lite $9.99, Plus $30 (10K cr), Pro $90 (40K cr), Ultra $95–$300/mo; free tier w/ watermark. API: Ray-2 ~$0.08/s, Photon $0.015, Photon Flash $0.004. Enterprise provisioned throughput $2,100–$3,800/mo. Annual ~20% off.$1.06B total across 5 rounds. Series B $43M (Jan 2024, a16z). Series C $900M (Nov 2025, led by HUMAIN/Saudi PIF; co: a16z, Amplify, Matrix, AMD Ventures) at ~$4B valuation. Deal includes access to HUMAIN's 2-GW Project Halo supercluster (2026–2029). 1M users in 4 days at launch (Jun 2024, zero marketing); 25M registered users (2025/26); 500K DAU, 4M prompts/day. ARR ~$21.2M (Latka est —). CNBC Disruptor 50 #35 (May 2026). Adobe + Amazon Bedrock partnerships. Dream Lab LA = free filmmaker R&D space. [MIXED — users revenue ESTIMATE]2026-08-30
    MartiniMartiniCollaborative AI filmmaking workspace: directors compose shots via virtual camera position/lens/movement (not just text prompts) → generate via any of 21+ models (Veo 3, Kling, Sora, Flux, Seedance, MiniMax...) in one canvas → assemble a timeline → export XML to Premiere/DaVinci.Credit currency 'olives' (1 = $0.10). Free Indie (all models, public projects). Pro $30/mo ($30 credits, private). Studio $150/mo (team). Enterprise custom. Per-model per-second: Kling/MiniMax $0.10/s, Veo 3 Fast $0.20/s, Moonvalley $0.40/s, Veo 3 $0.50/s, Sora 2 Pro $0.65/s. No per-seat fees.YC W26 (~$500K standard). DG Daiwa Ventures also listed. No other announced rounds as of Aug 2026. Valuation n/a. 2-person team. 200+ films made in beta (founder-stated); films aired as TV commercials, in galleries, viral on social (founder-stated, unverified externally). X @martini_film 917 / @koh_terai 313 (scraped Aug 2026). Koh Terai spoke at NAB Show 2026 + Aspen Design 2025. Desktop app launched 20 Aug 2026.2026-08-30
    MiniMaxHailuo / MiniMaxMultimodal foundation-model lab shipping via consumer apps (Hailuo AI video/image/music; Talkie for US; Xingye for China) + open API. hailuoai.video generates video/image/music from text or image on web, iOS, Android globally.Consumer: Free (~3–5 videos/day); paid $7.99–$199.99/mo (5 tiers, ~40% off annual). Base 512p/6s = 12 cr; 1080p cinematic = 50–80 cr. API H3 per-second: $0.08/s @768p, $0.13/s @2K (10s 2K ≈ $1.30). Legacy: $0.28/6s @768p, $0.49/6s @1080p.Pre-IPO ~$850M across 4 rounds (MiHoYo, Alibaba, Tencent, HongShan/Sequoia China, Hillhouse, IDG, Gaorong). IPO Jan 9 2026 HKEX raised $619M at HK$165; closed HK$345 (+109%) → ~$13.7B mkt cap. Post-IPO ARR >$150M by Feb 2026. Ant Group lawsuit (~$433.5M) reported pending [UNCONFIRMED]. 236M+ cumulative users across 200+ countries; 214,000+ enterprise customers/developers; 600M+ videos generated (FY2025). Talkie ~11M US MAU (Jul 2024). FY2025 revenue $79M (+159% YoY), gross margin 25.4% (up from 12.2%), still loss-making ($251M adj net loss). H3 open-weight release Aug 2026.2026-08-30
    Mochi (MOCHI.TV)Mochi (MOCHI.TV)YC S26Mobile app to watch and create 1-minute vertical AI anime microdramas.Freemium app (coins/subscription); amounts not capturedYC $125K (Jun 2026) per third party; other rounds not disclosed2026-10-14
    MoonvalleyMoonvalleyBuilds + sells Marey — a text/image-to-video foundational model trained exclusively on licensed/public-domain footage, for pro filmmakers/studios/brands. Also runs Asteria Labs (in-house studio, acquired from Bryn Mooser & Natasha Lyonne) to deploy Marey on real productions.Consumer: $14.99/mo (100 cr), $34.99/mo (250 cr), $149.99/mo (1,000 cr); ~1–2 cr per 5s 1080p clip (inferred). Enterprise: custom models on proprietary IP (undisclosed).$154M total raised (seed + extensions 2023–Jul 2025): $70M seed (Nov 2024, General Catalyst + Khosla + Bessemer); ~$43M extension (Apr 2025); $84M seed extension (Jul 2025, GC lead; CAA, CoreWeave, Comcast strategic). Valuation not disclosed. 12+ large studios in paid pilots; ad agencies piloting; director Ángel Manuel Soto reported 20–40% production cost cut; Asteria using Marey for a Carl Sagan doc restoration. IG ~13K; X @moonvalley ~2,942 (approx); LinkedIn ~1,483.2026-08-30
    NarrativeNarrativeYC F25'Cursor for professional video editors': b-roll, search, selects, stringouts.—YC standard; other rounds not disclosed2026-10-14
    NativeNativeYC F26Autonomous marketing platform; 'Norn' model predicts which content ideas will outperform, then makes and posts them.—YC standard; other rounds not disclosed2026-10-14
    OpenAISoraText-to-video and image-to-video generation. Sora 2 (Sept 2025) added natively synchronized audio (dialogue, SFX, ambience) plus a TikTok-style social app with a vertical feed and consented, insertable AI likenesses ('Cameos').Consumer (now defunct): free at launch (free video removed Jan 2026); ChatGPT Plus $20/mo (limited); Pro $200/mo for Sora 2 Pro. API (sunsetting Sept 24 2026): Sora 2 $0.10/s (720p); Sora 2 Pro $0.30/$0.50/$0.70 per sec at 720p/1024p/1080p; ~50% batch discount.No Sora-specific raise. OpenAI: ~$40B round at $300B valuation (Mar 2025); $6.6B employee secondary at $500B valuation (Oct 2025) — world's most valuable startup. Disney's reported ~$1B stake + partnership collapsed by Mar 2026. #1 US App Store in 48h (164K dl); 1M downloads in <5 days (faster than ChatGPT). Peaked ~3.3M dl Nov 2025 → ~1.1M by Feb 2026. DISCONTINUED: app shut Apr 26 2026, API ends Sept 24 2026. Kill reason: ~$1M/day opex (WSJ) vs ~$2.1M lifetime in-app revenue; copyright/deepfake backlash (Disney block, SAG-AFTRA).2026-08-30
    OpenArtOpenArtAll-in-one AI visual storytelling platform: text prompt/reference image → consistent AI characters → images → video clips → full 1-min+ stories with music, audio, lip-sync, export to social. Also Director (multi-scene long-form), 3D world gen, full image editor.Freemium credit-based tiers (free w/ daily credits; paid ~$12/mo up to ~$240/mo Wonder tier); teams $35/mo per member; annual saves 27–50%; extra-credit add-on $15/mo.$35M total: $5M seed (2023, Basis Set + DCM) + $30M Series A (30 Jan 2026, Canaan lead + Basis Set + DCM). Profitable before the Series A. Valuation not disclosed. $70M+ ARR (Jan 2026, 7x from $10.5M in 2025); 10M+ creators claimed, 8M MAU at Series A; ~20-person team ($3.5M rev/employee). Discord 350K+; IG ~368K; X ~72K. 6M users made videos within weeks of One-Click Story launch.2026-08-30
    PalettePaletteYC S26AI-native media engine: one conversational platform plus API to generate, edit and template multimodal content.Individual $0.01/credit (1,000 credits ~ a full character reel)YC standard; other rounds not disclosed2026-10-14
    PikaPikaAI video generation (text/image/video-to-video) aimed at social-media creators making short-form vertical content. Web + iOS; API via Fal.ai.Free (80 cr/mo, 480p, watermark, no commercial) · Standard $8/mo · Pro $28/mo · Fancy $76/mo (annual). Burn: I2V $0.04/s @720p, $0.06/s @1080p; Pikaffects $0.05/req.~$135M raised: Series A $35M (Lightspeed, Nov 2023), Series B $80M (Spark Capital, Jun 2024) at $470M valuation. Angels: Jared Leto, Adam D'Angelo, Nat Friedman, SV Angel. a16z is NOT an investor. 16.4M registered users; 1.4M avg MAU (H1 2025); 500K+ users in first 6 mo via Discord beta. Revenue undisclosed; 3rd-party ARR estimates $7.6M–$130M+. [users — Fortune Oct 2025]2026-08-30
    Pixley AIPixley AIYC F25Kids' drawings and ideas become personalised animated cartoon episodes guided by parents' lessons.—YC standard; other rounds not disclosed2026-10-14
    Pollo AIPollo AIOne-stop AI video + image platform aggregating 300+ frontier models (Kling, Runway, Veo 3, Hailuo, Sora, Seedance, Pika, Luma...) under one credit subscription. Text/image-to-video, style transfer, talking avatars, AI animation + developer API (api.pollo.ai, Jul 2026).Free (20 cr/mo, watermark) → Lite $10/mo (300 cr) → Pro $25/mo (800 cr, camera control) → Pro+ $50+/mo (2K cr) → Agency $100–$200+/mo (up to 50K cr). Annual ~14% off; commercial rights on paid. API USD, per-model rates unlisted.$14M seed (7 Dec 2025) — first institutional round, raised after profitability. Lead Gaorong Capital; co-investor ZhenFund. Valuation not disclosed. 20M+ registered, 6M+ MAU, 200K+ DAU (Dec 2025); $20M+ annualized revenue, break-even May 2025. Forbes Asia 100 To Watch 2026. Web ~6.9–8.1M visits/mo. X @itsPolloAI 110,610 (verified). Top markets: Japan, USA, India, Brazil.2026-08-30
    RemixRemixYC W26Auto-generates posts, carousels and short videos from your photos, notes and profiles with parallel agents.—YC standard; other rounds not disclosed2026-10-14
    ReplicateReplicateCloud API to run 50,000+ open-source & proprietary ML models with one line of code — no GPU/infra setup. Also fine-tuning, model sharing, community marketplace. Now in Cloudflare's Developer Platform (edge inference via Workers AI on roadmap).Pay-per-second GPU: T4 $0.000225/s ($0.81/hr), L40S $0.000975/s, A100-80GB $0.0014/s, H100 $0.001525/s ($5.49/hr), up to 8xH100 $0.0122/s. Some models per input/output. Volume + enterprise tier; no disclosed free tier.~$58M total: seed ~$5.3M (YC W'20 + angels Dylan Field, Guillermo Rauch), Series A $17.8M (Feb 2023, a16z/YC/Sequoia), Series B $40M (Dec 2023, a16z lead) at $350M post. Acquired by Cloudflare Nov 2025, terms undisclosed. 2M+ registered developers, 30,000 paying customers (Dec 2023); 50,000+ models by acquisition (Nov 2025). Customers: Unsplash, BuzzFeed, Character AI, Labelbox; indie: Pieter Levels' PhotoAI, Danny Postma's HeadshotPro ($1M+ ARR). Revenue ~$5.3M (2024, Tracxn est — unverified). Cog 9,500 GitHub stars. [MIXED — counts HIGH, revenue LOW]2026-08-30
    RunwayRunwayResearch company building "world models"; monetizes generative video for film/media/ads ("a new kind of camera"); hosts rival models + Media RouterSubscription + API + enterprise/studio deals; tier prices not published$5.3B (Series E, $315M, General Atlantic, Feb 2026); prior ~$3B (Series D, Apr 2025); ~$1.05B total raised ~$90M ARR Jun 2025 → $265–300M est. end-2025 (ESTIMATE, not officially disclosed; Sacra/getlatka); 2024 recognized rev $44M (+236% YoY, ESTIMATE). ~$155M 2024 EBITDA loss — not profitable2026-08-20
    Shengshu TechnologyViduFrontier Chinese AI video model + platform — text-to-video, image-to-video, and its signature reference-to-video; now pivoting toward a 'general world model.' Distributed via web app, Vidu API (MaaS), and Alibaba Cloud Model Studio.Free (800 cr); Standard $8/mo (commercial rights, no watermark); Premium $28/mo (4,000 cr); Ultimate $79/mo (up to 8,000 cr); Enterprise custom. API credits ~$0.005 each; Q3 one-time packs ~$19.90. Aggressively cheap entry vs Western rivals.Angel ~¥100M (~$14M, Jun 2023, Ant + Baidu); Series A+ >¥600M (~$82M, Feb 2026); Series B ¥2B (~$290M) led by Alibaba Cloud, Apr 2026 (+ TAL, China Internet Investment Fund, Baidu Ventures). ~¥2.6B (~$372M) raised in 2026. Valuation undisclosed. Note: Alibaba both funds and competes with Vidu. 10M+ users across 200+ countries (cumulative); claims >10x user & revenue growth in 2025 and works with '90%+ of film/entertainment stakeholders'; Q3 claimed #1 on Artificial Analysis + SuperCLUE reference-to-video leaderboard. [users; benchmarks vendor-sourced]2026-08-30
    ShofoShofoYC W26Index of billions of videos; agents build custom labelled video datasets for labs.Not publicYC standard; other rounds not disclosed2026-10-14
    shortkitshortkitYC W26SDK for any app to ship a TikTok-grade vertical video feed.Not publicYC standard; other rounds not disclosed2026-10-14
    SunoSunoAI music-creation platform — text/hum/voice prompt → full song (lyrics, vocals, instrumentation) in seconds; consumer, not pro toolConsumer subscription (freemium); tier prices not published$5.4B (Series D, Jun 2026, Bond/press); ladder $500M→$2.45B→$5.4B ~$300M ARR (Feb 2026, 2M paid subs X); ~$200M recognized revenue Nov 2025; trajectory $50M→$140M→$300M. Revenue ~100% subscriptions2026-08-20
    SynthesiaSynthesiaEnterprise AI avatar video platform — type a script, pick or clone an avatar, generate studio-grade talking (now full-body) video in 160+ languages. Deliberately enterprise-first (training, L&D, internal comms). Rebranded to an 'AI video communications platform for the future of work' (2.0/3.0) + dubbing, doc-to-video, agentic avatars.Freemium SaaS. Free tier; Starter ~$29/mo ($18 annual); Creator ~$89/mo ($64 annual); Enterprise custom (low-five-figures+/yr). Priced by video minutes, avatars, seats.~$535M+ raised Seed→Series E. Series C (2023, $90M, Accel + NVIDIA NVentures) = unicorn/$1B; Series D (Jan 2025, $180M) at $2.1B; Adobe strategic (~£10M, Apr 2025); Series E $200M at $4B post-money (Jan 26 2026, led by GV + NVentures/Accel/Kleiner/NEA). Reportedly rejected a ~$3B acquisition offer. Crossed $100M ARR Apr 2025 (official); ~$146M by Sept 2025, tracking toward ~$200M in 2026. 90% of Fortune 100, ~70% FTSE 100, 65,000+ customers, 1M+ users, 150k+ personal avatars (Bosch, Merck, SAP, Microsoft, Amazon). NRR >140%. [ARR Apr2025; later figures]2026-08-30
    tday.comtday.comYC P26 (Spring 2026)Turns every code release into on-brand launch videos, showcases and graphics, auto-published.Free $0 (100 credits); Pro $19/mo; Growth $149/mo; Unlimited $499/moYC standard; other rounds not disclosed2026-10-14
    Tell if AITell if AIYC F25Agents on a fleet of real phones create and post video content across hundreds of accounts, rev-share model.Revenue shareYC standard; other rounds not disclosed2026-10-14
    TryNearbyTryNearbyYC S26Subscription that sends local creators to film and post reviews for restaurants.$399/monthYC standard; other rounds not disclosed2026-10-14
    VoyagerVoyagerYC F26Open, local 'Codex for creative work': an LLM harness that steers 100+ image/video models and desktop tools.Free tier 1,000 credits/month or bring your own Claude/ChatGPT; paid tiers not capturedYC standard; other rounds not disclosed2026-10-14
    VyraVyraYC F26Autonomous video editing for non-editors: upload footage, describe the edit, get a polished cut.MCP Starter $24/mo, MCP Pro $65/mo; Vyra AI Starter $54/mo, Pro $129/moYC standard; other rounds not disclosed2026-10-14
    WideframeWideframeYC W26AI coworker for video editors: search, label, organise, sequence footage outside the NLE.$100/month, 7-day free trialYC standard; other rounds not disclosed2026-10-14
    YouArtYouArtYC P26 (Spring 2026)'Patreon for AI originals': storytellers create AI films/series with a video agent and get funded by fans.—YC standard; other rounds not disclosed2026-10-14