At his GTC 2026 keynote in San Jose on Monday, Nvidia CEO Jensen Huang declared "the ChatGPT moment for autonomous driving has arrived" — and backed the claim with a major new partnership with Uber.

Nvidia + Uber: L4 Robotaxis at Scale

Under the agreement, Uber will deploy Level 4 robotaxis powered by Nvidia's Drive AV software across its ride-hailing network. The rollout begins in Los Angeles and San Francisco in 2027, with plans to expand to 28 cities across four continents by 2028.

The vehicles will run Nvidia's Alpamayo 1.5 model — a reasoning vision-language-action (VLA) model that ingests driving video, ego-motion history, navigation guidance, and natural language prompts to generate driving trajectories. Nvidia says the model makes it substantially easier for vehicles to handle unpredictable road events, adverse weather, and complex pedestrian behavior.

New Auto Partners

Huang also announced four new automakers joining Nvidia's DRIVE Hyperion platform: BYD, Hyundai, Nissan, and Geely. These manufacturers will build next-generation driver assistance and autonomous systems on Nvidia's full-stack AV software — a notable expansion into the global auto market.

New Physical AI Models

Beyond automotive, Nvidia released Cosmos 3 for generating synthetic training worlds, and Isaac GR00T N1.7 — an open reasoning VLA model for humanoid robots that Nvidia says is now commercially viable for real-world deployment.

Nvidia's aggressive push into physical AI signals its ambition to own the software layer of autonomous transport — not just supply the silicon underneath it.