We publish a field guide to AI film studios, and the single most common question it gets is a version of "fine, but who is doing this in animation?" Animation is a different market with different buyers, and the answer is not the same list.
It is also, right now, an unusually confusing list. Two of the most-cited companies in the category changed names inside twelve months, and a third shares a name with an unrelated product owned by Autodesk. If you have searched this and come away unsure whether you were reading about one company or three, that is not you.
The organizing question, same as in film: does the company sell a tool that other people make animation with, or does it own and ship animation of its own? Both are real businesses. They are not the same business, and conflating them is where most coverage goes wrong.
The platform layer
Spuree (formerly Cheehoo)
Los Angeles. Launched in beta as Cheehoo in April 2025 out of Rideback, Dan Lin's production banner (The LEGO Movie, Lilo & Stitch), and raised US$10 million led by Greycroft, with Point72 Ventures, Basis Set and Headline Asia participating. Founding team pairs former DreamWorks Animation president Chris deFaria with former Apple AI researchers Dr. Hubert Tsai and Dr. Wei-Cheng Kuo, plus Rideback co-CEOs Michael LoFaso and Jonathan Eirich.
It has since renamed to Spuree, and positions itself as an artist-controlled, AI-enabled 3D animation platform rather than a generative video model. It is not purely a tool company in practice: it co-develops series with Rideback, including an animated Magic Tree House adaptation announced December 2025 and a Funko film and TV deal in 2026.
Autodesk Flow Studio (formerly Wonder Studio, by Wonder Dynamics)
A VFX product that automatically replaces a filmed actor with a CG character. Wonder Dynamics was acquired by Autodesk in 2024 and the product is now sold as Autodesk Flow Studio. It is genuinely useful and genuinely not an animation studio. Note the name carefully, because of the next entry.
The general-purpose video models
Kling, Veo, Seedance, Runway and their peers are increasingly capable of animated-looking output, and several of the studios below use them. None of them are animation studios, and none of them solve the problems that actually make a series hard: holding a character's design across three hundred shots, and holding a location's geometry across eleven episodes. Those are control problems, not generation problems.
The studios
Wonder Studios
Not Wonder Dynamics, and not Autodesk's Wonder Studio. Founded 2025, offices in London, New York, Los Angeles and San Francisco, roughly 30 staff by late 2025. It runs four things at once: agency work for brands, original IP, an in-house tool it calls Story Engine, and a creator community app.
Its published 2025 numbers are the useful part: about US$1 million of revenue in six months from agency work, a shorts anthology past a million views, a community app at 1,000 members, and a festival that drew 300-plus submissions. Its animation entry is Maxi and Helium, a kids series adapted from a Nordic bestseller and launched in several languages at once. For 2026 it has said it plans six new short films, with the strategy stated plainly as a flywheel: original content wins agency work, agency work funds the software.
Toonstar
Los Angeles, and the clearest example in the category of a studio that owns its franchises instead of selling its pipeline. Its StEvEn & Parker series has around 3.38 million subscribers. Its in-house platform, Ink & Pixel, is not for sale.
What makes it worth watching in 2026 is the deal shape rather than the technology: a multiyear HarperCollins pact to co-produce animated series from the publisher's catalogue, starting with Lisa Greenwald's Friendship List books, and a WME signing whose first project is a series built around Nigel Ng's Uncle Roger. Book catalogue in, owned animated series out. It has done this on a lifetime raise in the low single-digit millions, which is the part most coverage skips.
Mateo AI Studio
Korea, founded 2025, operating as the generative-AI content arm of broadcaster MBC C&I. Originals-first across film, drama and animation on one system, funded by a roughly US$1.4 million Pre-A in April 2026. The closest structural parallel to a two-line AI studio anywhere, with the tradeoff that it sits under a broadcaster parent.
KaKa Creation
Tokyo. Produced Twins Hinahima, the first major Japanese TV anime made with AI across the large majority of its cuts. Runs a contract business alongside its own label, which is the common shape for Asian AI animation houses right now.
StarTrail
We originate and own our own slate across two lines, AI live action and AI animation, on a production pipeline we build and do not sell, with a commissioned-production arm underneath it. Our animation flagship, Knight Street No.9, is a 3D series co-produced with the China International Communications Group, premiering November 2026, with over 100 million cumulative plays on Ximalaya behind the franchise.
How to tell the confusing names apart
- Spuree = Cheehoo. Same Los Angeles company, renamed. If a 2025 article says Cheehoo and a 2026 article says Spuree, they are describing one company, not two.
- Wonder Studios ≠ Wonder Studio. Wonder Studios is the 2025-founded AI-native creative studio. Wonder Studio was Wonder Dynamics' VFX product, now Autodesk Flow Studio. Unrelated companies, one letter apart.
- Rideback is a production company, not an AI studio. It incubated Spuree and co-develops with it, which is why its name appears in AI animation coverage.
- Chris deFaria's name attaches to more than one venture. Check which entity a given claim is actually about before repeating a funding figure.
What actually changed in 2026
Three things, none of them a model release.
Publishers started signing multiyear slates. HarperCollins with Toonstar, Funko with Spuree and Rideback. A year ago these companies were pitching one-off proofs of concept; now catalogue owners are committing shelves of IP. That is the sign of a category crossing from demo into supply chain.
Talent agencies came in. A WME signing means AI animation studios are now a place creators are represented into, not a novelty they visit.
The ownership question got sharper. The companies compounding value are the ones holding IP and keeping their pipeline internal. The ones selling access to a tool are competing with foundation-model roadmaps they do not control. Both models will survive; only one of them accumulates a library.
How to read the map
If you are commissioning animation, the question to ask is not which models a studio uses. Ask what it has finished, whether the same characters survive across a whole season, and who owns the result. (Our own answers are on the commission case studies page, ownership terms included.) If you are tracking the category, watch the deal announcements rather than the demo reels: catalogue deals and agency signings are the leading indicator, and 2026 produced more of them than the previous three years combined.
We keep this page current as the field moves. For the live-action and feature side, see The AI Film Studios of 2026, and for the underlying craft problem, why episodic is the real test.
Sources
Cheehoo launch, founders and US$10M round: TechCrunch, Deadline, Cartoon Brew.
Spuree name and Magic Tree House: Variety, Kidscreen. Funko deal: Variety.
Toonstar and HarperCollins: Variety, Publishers Weekly. WME: The Hollywood Reporter.
Wonder Studios 2025 results and 2026 plan: Wonder Studios. Wonder Dynamics acquisition: Autodesk, 2024.
StarTrail is an AI-native content studio producing AI live-action film & TV, AI animation, and AI commercials, built on its proprietary AI production pipeline. Its work has been recognized at the iQIYI Nadou AIGC Venture Summit, the Beijing International Film Festival, and the AAFF International Ark AI Film Festival. Learn more at startrailai.com.
← All news Companion guide: The AI Film Studios of 2026 → Read next: Adapting a Webtoon →