# Decart's world model is real — and narrow on purpose

> Decart's Oasis 3 (10 June 2026) is a real-time driving world model via API — focused, not universal.

*It dropped the sci-fi 'simulate anything' pitch for something that might actually sell.*

By The InsidersFeed Desk · InsidersFeed
Canonical: https://insidersfeed.com/news/decart-oasis-3-narrow-on-purpose

> **Key:** **The take:** the impressive thing about Oasis 3 isn't that it's flashier — it's that Decart got *less* ambitious in the right way. The old Oasis was a 'simulate any world' demo. This is a driving simulator with an actual customer. That's progress, not a downgrade.

Launched 10 June: a world model that generates photorealistic, promptable driving scenes in real time, multi-camera, via API at ~$0.02/sec. Target buyer: AV teams who need to rehearse rare, dangerous scenarios at scale. CEO Dean Leitersdorf wants developers to build on it like the early OpenAI ecosystem — platform ambition under a focused product.

## The caveat that actually counts

Outlets that tried it flagged real-world flaws. Normally 'demo has rough edges' is forgivable. Here it's the whole ballgame: if you're training a self-driving car against a simulation, the *gaps* between sim and reality are precisely where cars crash. A world model that's 95% photorealistic can teach a 5% wrong lesson, and you find out on a real road.

> **Note:** **Credit:** Decart shipped via API on day one (developers can actually use it), raised ~$300m at ~$4bn with Toyota and Nvidia in, and picked a domain where simulation already has a real budget. This isn't vapourware — it's an early product in a hard category.

So: genuinely interesting, appropriately scoped, and don't let 'photorealistic' do too much work. The bet that pays off isn't the pretty driving clip — it's whether developers build enough on top that Decart becomes the layer everyone else's physical-AI uses. Long road. Real start.

## FAQ

### Is Oasis 3 good enough to train self-driving cars?
It's a real, usable tool for generating rare driving scenarios at scale — but testers flagged real-world flaws, and sim-to-reality gaps are exactly where AV training is risky. It's best seen as a powerful complement to real-world testing, not a replacement.

### Why did Decart narrow Oasis to driving?
Its earlier versions were open-ended 'explore any world' demos. Focusing on driving trades breadth for domain-specific accuracy and, crucially, gives it a paying customer base (AV companies) — a more commercial, more defensible bet.