news
A studio just released a 135-minute, fully AI-generated Homer adaptation, made entirely with Kling
Fountain 0's Odysseus: The Fall is being pitched as the first AI-generated feature at Hollywood-blockbuster scale, timed to land days before Christopher Nolan's $250 million Odyssey.
In July, the AI film studio Fountain 0 released Odysseus: The Fall, a 135-minute adaptation of Homer's Odyssey built entirely with Kling. It is the most ambitious claim yet in AI filmmaking: not a proof-of-concept short, not a trailer, but a feature-length release the studio is pitching at the scale of a Hollywood blockbuster.
The director is Ash Koosha, who spent the months prior finishing Dreams of Violets, an AI-generated docu-drama about the 2022 Iranian protests that premiered at the Tribeca Festival. Koosha reportedly worked on Odysseus in parallel, spending around three months on a film centred on a hero he says he idolised as a child.
No script, just notes
According to reporting on the project, Odysseus was built from notes rather than a traditional screenplay. That production process has no real equivalent in conventional filmmaking, where a locked script normally precedes anything expensive being generated. That workflow is itself part of the story: it is only viable because generation is cheap enough to iterate on directly, rather than plan around.
The film leans into the tragic reading of the Odyssey rather than the heroic one. Koosha has described it as centred on “the fractured memory of a drowning man in his final minutes,” a version of the story that treats the voyage as trial rather than triumph.
The timing is not an accident
Odysseus: The Fall went up for rent and purchase on Fountain 0's own site days before Christopher Nolan's The Odyssey (a conventionally-shot, reported $250 million theatrical production) opened on 17 July. Being the AI-generated shadow release of a major studio tentpole, on the same source material, in the same month, is a marketing choice as much as a creative one, and most of the coverage treated it that way.
Why this matters more than another demo reel
Short AI-generated clips proving a model can produce a coherent 8-second shot are common; a 135-minute feature that holds together is a different claim entirely, because it requires consistency (of character, of world, of tone) across a runtime that punishes drift. Whether Odysseus actually delivers that consistency is a separate question from whether it exists at all. We have not watched it end to end and are not reviewing it here; we are reporting that it exists, at this length, on this model, because that alone is a new data point for what Kling can be pushed to do outside a fixed benchmark prompt.
It also lands squarely inside our own coverage: Kling is one of the three models in our benchmark round, where it scored well on multi-shot consistency but generated zero audio across every lip-sync prompt we ran. A 135-minute narrative feature is exactly the kind of use case where that audio gap would either have been engineered around or would show.
Sources: The Hollywood Reporter, Variety, Forbes.
Related: Kling 3.0
Daniel Ochoa: Covers model launches, shutdowns and pricing changes as they happen. Reads deprecation notices for a living so you do not have to.