Xpeng Kicks Off Mass Production of L4 Robotaxi, Targets Fully Driverless Operations in 2027

Xpeng has started mass-producing its Level 4 robotaxi, which uniquely relies on cameras alone, ditching LiDAR and HD maps, and is powered by its in-house Turing AI chips.

This move positions Xpeng to compete directly with established players like Baidu and Pony.ai by offering a scalable, cost-effective manufacturing platform, shifting the race from software demos to mass production.

The start of production was made possible by a convergence of factors: timely regulatory approvals from Guangzhou, the maturation of its autonomous driving software, and growing market demand validated by competitors.

Xpeng has officially started mass production of its first Level 4 robotaxi, marking a significant step in the autonomous vehicle race.

This isn't just another vehicle rolling off the assembly line; it represents a bold strategic bet. Xpeng is pursuing a 'pure-vision' approach, similar to Tesla. This means the robotaxi navigates using only cameras and its powerful in-house 'Turing' AI chips, without relying on expensive LiDAR sensors or constantly updated high-definition (HD) maps. The core idea is that this method is more scalable and cost-effective, allowing for faster expansion across different cities and regions compared to map-dependent systems.

So, why is this happening now? The decision is rooted in a clear causal chain. First, recent developments created the perfect window of opportunity. The Guangzhou government granted Xpeng the necessary permits for L4 testing on public roads in early 2026. This regulatory green light coincided with the maturation of Xpeng's core software, VLA 2.0. Furthermore, a recent service outage involving a competitor's robotaxis in Wuhan highlighted the importance of system reliability, allowing Xpeng to frame its 'aviation-grade' redundant design as a key advantage.

Second, the market groundwork was already laid by competitors. Companies like Baidu's Apollo Go have demonstrated that there is significant public demand for driverless ride-hailing services, with weekly rides reaching into the hundreds of thousands. This validation of the business model reduced the risk for Xpeng and provided a clear incentive to enter the market with a scalable solution.

Third, this move is the culmination of a long-term strategy. At its 2025 AI Day, Xpeng laid out this exact roadmap: a robotaxi built on a pure-vision stack with 3,000 TOPS of computing power. This long-term planning, supported by a favorable policy environment and strong investor belief in the sector's potential, has now transitioned from a blueprint to a physical product.

Beyond its own robotaxi fleet, Xpeng is also positioning itself as a technology supplier. The company plans to offer its Turing chips and VLA 2.0 software to other automakers, with Volkswagen already signed on as a partner. This creates an additional, high-margin revenue stream beyond vehicle sales and ride-hailing fees.

In essence, Xpeng's announcement shifts the narrative from 'if' autonomous vehicles will be commercialized to 'how fast' they can be scaled. By industrializing a cost-effective, vision-only L4 system, Xpeng is not just joining the robotaxi race—it's trying to change the rules.

L4 (Level 4) Autonomous Driving: A classification for vehicles that can perform all driving functions under specific conditions (e.g., within a geofenced urban area) without any need for a human driver to intervene.
Pure-vision stack: An autonomous driving system that relies exclusively on cameras to 'see' and interpret the world. It does not use other sensors like LiDAR or pre-built high-definition maps.
TOPS (Tera Operations Per Second): A measure of an AI chip's processing power. One TOPS equals one trillion calculations per second, indicating how quickly the chip can run complex AI models.

Xpeng Kicks Off Mass Production of L4 Robotaxi, Targets Fully Driverless Operations in 2027

Xpeng has started mass-producing its Level 4 robotaxi, which uniquely relies on cameras alone, ditching LiDAR and HD maps, and is powered by its in-house Turing AI chips.

Xpeng has officially started mass production of its first Level 4 robotaxi, marking a significant step in the autonomous vehicle race.

L4 (Level 4) Autonomous Driving: A classification for vehicles that can perform all driving functions under specific conditions (e.g., within a geofenced urban area) without any need for a human driver to intervene.
Pure-vision stack: An autonomous driving system that relies exclusively on cameras to 'see' and interpret the world. It does not use other sensors like LiDAR or pre-built high-definition maps.
TOPS (Tera Operations Per Second): A measure of an AI chip's processing power. One TOPS equals one trillion calculations per second, indicating how quickly the chip can run complex AI models.