QCraft Opens European Headquarters in Germany
Published on
5 min read

QCraft Opens European Headquarters in Germany, Partners with Qualcomm

Chinese autonomous driving solutions maker QCraft has opened its European headquarters in Germany. According to Reuters, QCraft made the announcement at the IAA Mobility conference. Following QCraft’s opening of European headquarters in Germany, delivery of the company’s full-stack solutions in the U.S., South Korea, and Japan is set to commence in 2026.

“These milestones mark QCraft’s official entry into the EU and global markets,” the company said in a statement.

Building on Qualcomm’s Snapdragon Ride Platform

QCraft also announced that it is partnering with U.S. chip maker Qualcomm to fast-track global deployment of its L2++ to L4 autonomous driving solutions. The difference between these levels of driver assistance solutions is human supervision. L2++, which is commonly known as Level 2 Plus, is more advanced than Level 2.

Level 2 is a partial automation standard featuring enhanced capabilities such as automated lane changes and hands-off driving in specific environments. Both levels require full driver supervision. In July 2025, Tesla’s assisted-driving software outperformed Chinese rivals in a Level 2 autonomous driver-assistance system test. Tesla’s software scored better than Xiaomi, BYD, and Huawei in tests conducted on highways and urban areas in China.

Level 4, on the other hand, features high-level automation where vehicles drive themselves completely in defined environments like cities or highways without requiring driver supervision.
The QCraft-Qualcomm partnership will also see the Chinese driving solutions maker build next-generation intelligent assisted driving solutions on Qualcomm’s Snapdragon Ride platform.

The latest collaboration will bring together QCraft’s technical capabilities and Qualcomm’s ecosystem in a way that reduces cost and production cycles while providing customized solutions that meet the needs of global automakers. QCraft has already completed development of its Navigate on Autopilot (NOA) solutions on Snapdragon Ride. Besides Qualcomm, QCraft also works with NVIDIA and Horizon Robotics chipsets.

The partnership announcement comes soon after Qualcomm and BMW launched the automated driving system, Snapdragon Ride Pilot, for hands-free highway driving on September 5, 2025. Ride Pilot is built on the Snapdragon Ride system-on-a-chip and was developed jointly by Qualcomm and BMW.

Leveraging Germany’s Strong Auto Legacy

QCraft’s expansion to Europe through the Germany HQ reflects the company’s long-term global expansion strategy. Germany’s lasting legacy in automotive innovation, rigorous technical standards, and comprehensive industrial ecosystem provide a conducive environment for QCraft to localize its driver-assistance solutions.

At its European office, the autonomous driving startup will run critical functions including certification, business development and research and development to ensure that its solutions adhere to the highest EU standards.

Following the launch of QCraft’s EU base in Germany, the autonomous driving startup plans to launch its NOA solutions in about a million vehicles. The Chinese firm has earned high scores from leading automakers and attained the best safety metrics in the industry.

QCraft’s Safety Priority

QCraft adopts a safety-first approach in the development and large-scale production of its autonomous driving solutions. The company is built on a safety-first philosophy that guides its approach to autonomous driving advancement and innovation.

QCraft’s focus on safety also guides the evolution of its assisted driving solutions, which enables it to deliver unmatched security for users. Its technology is also designed for compute efficiency, an aspect that allows its solutions to deliver high performance with less computing power.

QrCraft’s technical efficiency emanates from its comprehensive data toolchain and large-model capabilities developed in-house from the ground up. This mix of core strengths allows its algorithms, including its VLA, world models, and end-to-end models, to iterate at extremely high speeds.

Linda Hadley
Scroll to Top