Moonvalley Raises $53M, Alibaba Unleashes WAN2.1-VACE AI Video Suite

This week, generative video AI advanced on all fronts, with major startup capital, open source disruption, and billions gaining direct access to automated video creativity. COEY delivers the essential updates for May 10-16, 2025.

Moonvalley Grabs $53M to Challenge AI Video Titans

Santa Monica-based Moonvalley made headlines with a $53 million funding round announced May 16, as reported by TechCrunch. The company’s platform, known for script-driven, pro-grade AI moviemaking with an emphasis on creative control, enters the race for dominance against giants like Runway and upstarts like Hedra (who themselves finished a $32M raise this week). Investors cited Moonvalley’s strengths in advanced scene control, camera choreography, and ethical dataset practices as deciding factors.

The new capital follows an earlier $43M raise, spotlighting the rush of private money into AI video for enterprise, advertising, and film markets. As both Moonvalley and Hedra amass over $80M between them in weeks, U.S. startups are primed for rapid scaling and high-velocity product iteration. Industry insiders now anticipate frequent, competitive feature drops as Moonvalley seeks to redefine what’s possible with automated visual storytelling at professional scale.

Alibaba Opens WAN2.1-VACE: A Broad, Open-Source Video AI Leap

On May 15, Alibaba debuted WAN2.1-VACE, a powerful open-source AI video suite, drawing immediate global attention. Covered widely in industry press, including Artificial Intelligence News and TVBEurope, WAN2.1-VACE gives developers and creators composable tools for precise video editing, scene-level control, and seamless text-to-video synthesis. The open approach, complete with model weights and documentation, removes the paywall for advanced experimentation and makes Alibaba’s innovation accessible to R&D teams and media startups around the world.

Industry feedback highlights WAN2.1-VACE’s balance of fast inference, robust motion accuracy, and frame-precise customization. Early testing is underway across GitHub and Hugging Face. The toolkit is already positioned as a prime open alternative to proprietary video AIs in the West, like OpenAI’s Sora and Google’s Veo 2, and sets a new baseline for what enterprise and indie teams can deploy, fork, or extend in their own workflows. Researchers anticipate formal benchmarks, and a new round of competitive improvements, in the weeks ahead.

TikTok AI Alive Launches Worldwide: Automated Animation for All

May 13 marked the global rollout of TikTok’s AI Alive, a new feature seamlessly embedded in TikTok Stories. The tool enables any user to transform a static photo and a short prompt into a lively, AI-animated video clip, complete with blinking eyes, facial gestures, and synthesized speech. Major outlets including CNET and Android Authority are hailing the feature’s real-time speed and entertainment value, though some facial artifacts persist depending on image quality.

AI Alive democratizes creative video, letting everyday users participate in the generative revolution, no pro skills required. The launch instantly puts AI-powered animation tools in the hands of over a billion people and intensifies the race among big social platforms to unlock mass-market synthetic media. YouTube (Veo) and Meta are now under renewed pressure to respond with their own advanced generative video features.

HeyGen CTO: The Global AI Localization Challenge Grows

This week, in an interview with Slator, HeyGen CTO Rong Yan addressed the technical hurdles of making video AI work natively in dozens of languages. The company’s focus on dubbing accuracy, lip sync, and mouth movement across cultural spectrums is now a key driver for enterprise and broadcaster adoption. While HeyGen did not launch a new feature this week, its ongoing research underscores the sector-wide challenge of cross-lingual, culturally sensitive avatars, rapidly becoming a “must have” for international markets.

Rising Threat: Fake AI Video Sites Spread Infostealer Malware

The explosive demand for generative video has also triggered a surge in cybersecurity threats. According to reporting from Lifehacker, BleepingComputer, and Morphisec, a new wave of malicious "AI video generator" websites is distributing infostealer malware disguised as breakthrough generative tools. These sites specifically target creatives and developers seeking free downloads, harvesting browser credentials and sensitive data. Experts urge all users to download models and apps strictly from trusted sources like official vendor sites or GitHub, as the attack surface for fake tools grows alongside legitimate interest.

Open Source and Industry Pulse

• Alibaba’s WAN2.1-VACE is the only major open-source video model launch this week, now trending on developer repositories and social feeds worldwide.

• Lightricks , Israel’s vision AI team , continues work on model speed and efficiency, but has no new release post-May 10.

• No notable U.S. open-source launches have landed this week, but cross-platform compatibility is now a key research topic thanks to Alibaba’s release.

Funding and Market Dynamics

Forbes and TechCrunch report a continued scramble among AI video startups for top-tier venture backing, new data partnerships, and creator-centered API features. With Moonvalley and Hedra raising in record numbers, and open source tools from China gaining traction, the battle for talent and cutting-edge differentiation is driving the sector’s rapid evolution.

• Moonvalley cements its place as a top-funded, pro-focused generative video platform.

• Alibaba’s WAN2.1-VACE sets a new global open-source standard for video AI development and adoption.

• TikTok’s AI Alive makes advanced, user-friendly automated video animation an everyday reality for billions.

• HeyGen keeps localization and avatar realism in sharp focus for the global enterprise market.

• Security risks surge as opportunistic cybercriminals target the swelling pool of generative video creators and experimenters.

The generative video market this week is defined by velocity, accessibility, and scale: from $50M+ bet rounds in Los Angeles, to open source disruption out of Hangzhou, to everyday AI-powered animation on the world’s biggest social network. Stay tuned as the arms race moves into an even higher gear.