Einride will evaluate NVIDIA’s driving AI in trucks, initially without vehicle control

Einride will evaluate NVIDIA’s Alpamayo driving model inside its autonomous trucking system, initially without giving it vehicle control. In an October 21, 2026 announcement distributed by GlobeNewswire, Einride outlined a collaboration to adapt the AI to freight operations. A live road evaluation is planned for the second half of 2027.

The companies are targeting a practical development problem: extending the Einride Driver autonomous driving system to more vehicle types and routes. Einride says adapting Alpamayo to its vehicles, routes and operating data is intended to shorten development cycles and reduce that expansion’s cost. The wider objective is to scale autonomous service across highway and point-to-point routes within a freight network already serving customers.

Turning a general driving model toward freight

Alpamayo is a vision-language-action model, or VLA. In the companies’ description, it interprets a driving scene, proposes how the vehicle should move and explains its reasoning.

Einride will use its proprietary operational data for post-training and distillation of Alpamayo. The company presents data from its real freight operations as the ingredient needed to turn a general driving model into something suited to trucking.

A live comparison, not a handover

The initial testing approach is shadow mode: Alpamayo runs live, but its outputs do not control the truck. Einride says those outputs will be evaluated against its driving planner and human driver performance.

Einride also draws a boundary around the model’s intended place in the system. Alpamayo will operate within the Einride Driver, but Einride says its safety architecture will retain full authority over the vehicle. The announced role is therefore a reasoning model inside a controlled driving system, not a replacement for the company’s safety architecture.

An addition to the September agreement

The Alpamayo work extends a strategic relationship announced on September 21. Under that earlier agreement, Einride is building the next generation of the Einride Driver on NVIDIA’s Hyperion platform, integrated with NVIDIA’s Halos safety system. The relationship also includes training and validating driving models using NVIDIA Cosmos and Blackwell infrastructure.

NVIDIA’s automotive vice president, Rishi Dhall, described the new work as exploring whether reasoning-based AI can help trucks navigate complex driving conditions and accelerate development for new routes and vehicle types. His description ties the model to both Einride’s freight data and its safety engineering, rather than presenting Alpamayo alone as a finished trucking system.

Einride says it aims to be the first trucking company to deploy Alpamayo on Hyperion. That remains an ambition attached to this development program, not a completed deployment.