Beta Builder: What India's AI-Hybrid Film Means for Creators

Two films, one week, one lesson

On October 29, 2026, Abundantia Entertainment's AI division aiON and AI video platform InVideo began principal photography in Mumbai on Beta Builder, described as India's first generative AI hybrid feature film, according to Deadline and Indiantelevision. It's the first release under a five-film, Rs 100-crore ($10.5 million) co-production slate the two companies announced earlier in 2026.

In the same week, South Korea's fully AI-generated feature Steelay quietly set a different kind of record: at 90 minutes, it became the longest AI-made film yet, built by a team of just over 10 people over roughly six months using about 13,000 AI-generated video clips. One film leans on real actors and a real set. The other leans almost entirely on synthetic generation. Taken together, they're a useful natural experiment for any creator trying to decide how much of their own production to automate.

What Beta Builder is actually doing

Beta Builder is not a fully synthetic production. Per Filmofilia, the film combines live-action filming with AI-assisted production rather than dispensing with a shoot altogether. Hridaye A. Nagpal, InVideo's Global Creative Lead, is directing, and his own framing is the clearest summary available: 'Beta Builder is our bet on that future, with a live-action shoot and AI working side by side.'

The story itself is fitting for the moment: Jai, a self-taught architect from Dharavi, gets a shot at realizing his dreams inside a virtual reality world, only to learn that winning there means more than just proving himself technically. No release date, distributor, or streaming platform has been confirmed yet, and the brief is explicit that these details are still unknown. The wider five-film aiON slate is expected to start delivering projects sometime in 2027.

It's also worth noting the infrastructure behind this deal. InVideo has acquired GoBo Labs, launched India's first AI filmmaking lab with Whistling Woods, and now has this standing studio partnership with Abundantia. This isn't a one-off experiment; it's a company building repeatable pipelines for hybrid production, which is exactly why the workflow choices here matter more than the plot.

What Steelay reveals about the all-AI route

Steelay took the opposite bet: no conventional computer graphics, an effects-heavy sci-fi action story, and a production team small enough to fit in a single room. That's a genuinely compelling economics story for indie creators — a 90-minute feature from roughly 10 people in six months is not something traditional VFX pipelines can touch.

But the director, Seol Woo-in, flagged a workflow risk that every small AI-dependent production should take seriously: unpredictable updates to overseas AI tools. In Seol's words, 'a new version of the tool was released after we had already finished about 90 percent of the film, so we had to test it again and make additional revisions.' Tool updates or changes to credit/pricing policies hit schedule and budget directly, and Seol called this uncertainty the production's biggest challenge.

That's not a hypothetical problem for episodic short-form creators. If you're building a recurring AI cast or a multi-episode visual style, a mid-series model update can silently change character consistency, art style, or voice quality — forcing rework you didn't budget for.

The practical takeaway: hybrid isn't a compromise, it's a strategy

Neither production went fully synthetic, and neither went fully traditional. Beta Builder keeps live actors for emotional scenes and lets AI handle scale and VFX-heavy elements. Steelay went further toward full AI generation but still needed a human creative team making directorial choices on every one of those 13,000 clips. The common thread is that someone is still directing, writing character briefs, selecting takes, and making the calls that AI tools don't make on their own.

For short-form creators running AI-driven channels, that translates into a concrete production model: keep a human showrunner layer even when the visuals are AI-generated. That means writing tight character briefs, directing voice and tone explicitly, and doing final cut selection yourself rather than accepting the first generated output. The AI handles volume and repetitive visual work; you handle the judgment calls that make a character or story feel consistent across episodes.

Lock your tools before you lock your schedule

The Steelay experience is a direct warning: version-lock your AI tools mid-project wherever the platform allows it. If a tool update happens in the middle of a multi-episode series, you risk the same problem Seol described — retesting and revising work you thought was finished, late in production.

Practically, that means checking whether your platform lets you pin a model version for the duration of a project, keeping a written log of exactly which model and settings generated each asset, and building a buffer into your schedule for at least one unplanned revision pass. This is exactly the kind of production discipline tools like Episodo are built to support, by letting creators lock in a character brief and visual direction and then generate against it consistently rather than re-prompting from scratch every episode.

None of this is hype. Beta Builder has no confirmed release date yet, and we don't know how audiences will respond to either film. But the production choices themselves — hybrid over all-or-nothing, human direction preserved, tool versions respected — are lessons creators can apply right now, independent of how either film performs.