CATL and Avatr Signal China EV’s Next Big Leap

China’s EV industry delivered three important signals at once this week: smart driving is being tested in one of Beijing’s toughest real-world road challenges, CATL says its sodium-ion batteries will enter mass production in 2025, and cockpit supplier Yanfeng argues the next in-car battleground is no longer pure hardware features but scenario-driven experiences. Taken together, these developments show how the Chinese EV market is evolving beyond the old range-and-price debate into a broader competition around autonomous driving capability, battery chemistry, cold-weather usability, and intelligent cabin design.

Beijing’s “Devil Road” Challenge Puts Smart Driving Under Pressure

On May 30, Beijing will host an elite smart-driving challenge on a notoriously difficult urban route described as a “devil road” section. The defending benchmark is the Avatr 12, which scored 112.5 points in the opening ranking event on May 28.

Now eight challengers are lining up:

• Xiaomi YU7
• Zeekr 7X
• IM LS6 (Momenta)
• NIO ET7
• Li Auto L9
• Wey Lanshan (Yuanrong/DeepRoute.ai)
• Hongqi HS6 (Zhuoyu)
• Deepal L06 (Horizon Robotics)

The field is unusually broad:

• Price range: RMB 146,900 to RMB 536,000
• Compute power: 100 TOPS to 1,016 TOPS
• Vehicle length: roughly 4.8m to 5.2m

That spread matters because it tests one of the biggest claims in China’s EV market: whether advanced driver assistance systems are becoming democratized, or whether premium pricing still buys a clear real-world advantage.

Avatr 12’s Real Edge May Be Sensor Redundancy, Not Price

The D1EV source argues that the biggest gap between the Avatr 12 and its rivals is not sticker price or even raw compute, but sensing hardware.

Avatr 12 is reported to use:

• 4 lidar units
• A primary lidar with 896 lines
• Detection of 14 cm low obstacles at 120 meters

By contrast, among the eight challengers:

• Most use 1 lidar
• Lidar resolution is listed in the 126-150 line range
• Hongqi HS6 reportedly has no lidar

In difficult urban environments such as ultra-narrow roads, blind under-bridge approaches, mixed traffic, and complex turns, that extra perception redundancy could be decisive. This is a useful reminder that smart driving performance is still heavily shaped by system architecture, not just software branding.

Smart Driving Challenge: Key Comparison

More TOPS Does Not Automatically Mean Better ADAS

One of the most revealing points from the report is that compute power alone is proving a poor predictor of real-world results.

The challengers reportedly fall into three rough compute camps:

• 500+ TOPS: NIO ET7, Xiaomi YU7, Zeekr 7X, Li Auto L9
• 250-400 TOPS: IM LS6, Wey Lanshan, Deepal L06
• Lowest tier: Hongqi HS6

Yet previous competition data cited by D1EV suggests the market is already learning a hard lesson: high compute does not guarantee high scores. The article references the XPeng P7, said to have 2,250 TOPS, but still placing only third in an earlier event. That highlights a critical truth in Chinese autonomous driving development today:

• Algorithm efficiency matters more than headline silicon
• Sensor fusion quality can outweigh peak compute
• Edge-case tuning is becoming a stronger differentiator than specs on paper

For NIO ET7, the question is especially interesting. On paper, its 1,016 TOPS should offer a major advantage. In practice, prior event data reportedly showed efficiency-related deductions, suggesting that translating hardware into polished urban NOA performance remains a software challenge.

The Dark Horses: Xiaomi YU7, Wey Lanshan, and Deepal L06

Three challengers stand out for different reasons.

Xiaomi YU7

Xiaomi YU7 looks like one of the most dangerous rivals to Avatr 12. D1EV cites a Hefei event-adjusted score of 98.21, backed by 700 TOPS and Nvidia-based compute. With some test scenarios removed in Beijing, the YU7 may have an even better chance to convert strong hardware into a headline result.

Wey Lanshan

Wey Lanshan may be the safest bet for consistency. Its 91.46-point reference score is described as highly credible, with the Yuanrong/DeepRoute.ai solution praised for stability across multiple competitions. In a difficult route challenge, consistency can beat theoretical peak performance.

Deepal L06

Deepal L06 could be the true value story. It combines:

• An entry price of RMB 146,900
• A relatively compact 4,830 mm body
• A Horizon Robotics smart-driving stack
• A cited Hefei score of 96.88, good enough for runner-up there

If it performs near the Avatr 12 benchmark at less than half the price, it would become powerful evidence that China’s ADAS race is entering a phase of real technology equalization.

Vehicle Size Still Matters on Tight Urban Roads

The Beijing route reportedly includes narrow roads and blind spots under bridges, which means physical packaging still matters.

The smaller, more maneuverable group includes:

• Deepal L06: 4,830 mm
• Zeekr 7X: 4,825 mm

The largest vehicles include:

• Li Auto L9: 5,218 mm
• Wey Lanshan: 5,156 mm
• NIO ET7: 5,101 mm

That matters because long-wheelbase SUVs and sedans face harder turning circles, more pressure in narrow passing situations, and potentially larger blind zones. Still, Avatr 12’s own earlier success as a roughly 5-meter-class vehicle shows that strong smart driving calibration can offset some of the physical disadvantages.

CATL’s Sodium-Ion Battery Push Could Reshape the Entry EV Market

While automakers battle over ADAS bragging rights, CATL is targeting a different bottleneck: battery cost, cold-weather usability, and safety.

At the 2026 Equipment Power Forum, CATL chief scientist Wu Kai, also a member of the Chinese Academy of Engineering, said the company will bring a series of sodium-ion battery products into mass production in 2025.

That is significant because sodium-ion technology has long promised lower raw-material dependence than lithium-ion chemistry, but commercialization has lagged. CATL is now positioning it as a practical near-term product, not just a research project.

CATL Sodium-Ion Battery: Claimed Highlights

Just as importantly, CATL says the battery passed severe abuse tests including:

• Multi-face compression
• Nail penetration
• Electric drill penetration
• Saw cutting

And according to the source, the cells did so with no fire and no explosion.

Why Sodium-Ion Matters More Than the Headline Range Number

If CATL can mass-produce sodium-ion batteries at scale, the impact may be strongest in three segments:

1. Entry-level EVs and PHEVs

   • Lower material costs could improve margins or enable more aggressive pricing.

2. Northern China and cold-climate markets

   • The claimed low-temperature charging and discharge performance addresses one of EV adoption’s most persistent consumer concerns.

3. Commercial and fleet applications

   • A 10,000-cycle life, if validated in real use, would be especially attractive for high-utilization vehicles.

For the Chinese EV market, this is strategically important because battery innovation is no longer only about increasing range. It is increasingly about reducing dependence on lithium, stabilizing supply chains, and improving all-weather usability.

CATL also disclosed a longer-term focus on lithium-air batteries, which it described as a next-generation battleground with a theoretical energy density 5-10 times that of today’s lithium batteries. That remains a long-horizon technology, but it shows how aggressively Chinese battery leaders are trying to defend their lead beyond LFP and NMC.

Yanfeng Says the Smart Cabin Is Moving From Feature-Led to Scenario-Led

The third major signal comes from Yanfeng, one of China’s most important automotive suppliers. Speaking at the 2026 Future Auto Pioneer Conference, Lu Chuanhua, vice president of product strategy and ecosystem cooperation at Yanfeng International, said the future cockpit will shift from function-driven to scenario-driven development.

That sounds subtle, but it is a major industry pivot.

Yanfeng says it generated about RMB 150 billion in sales in 2025, operates 230+ global locations, and employs nearly 60,000 people. The company describes itself as a Tier 0.5 supplier for intelligent vehicle upper-body systems, signaling a role that is more integrated than a traditional component maker but broader than a single-system supplier.

Yanfeng’s “Four Freedoms” for the Future Cabin

Lu outlined four development directions:

• Freedom of space
• Freedom of interaction
• Freedom of comfort
• Freedom of health and safety

The most notable technical point is safety. Lu emphasized that “safety is the greatest luxury”, and said Yanfeng has developed a world-first safety protection mechanism for zero-gravity seats that can protect occupants in collision scenarios even at reclined seating angles. That is increasingly relevant as Chinese automakers push lounge-style cabins, executive seating, and immersive rear-seat experiences.

Huawei Partnerships Show How Fast the Cabin Stack Is Consolidating

Yanfeng also highlighted its collaborative model through its EPG innovation platform and pointed to co-development with Huawei’s Harmony Intelligent Mobility Alliance. According to the source, that cooperation has already delivered first-launch technologies on multiple “Five界” models, including features for the Maextro S800 and Luxeed D9, such as angel seats and zero-gravity wraparound systems.

This matters because the Chinese EV race is becoming more ecosystem-driven. Competitive advantage increasingly comes from how well automakers and suppliers integrate:

• ADAS hardware and software
• battery chemistry and thermal management
• cabin electronics and operating systems
• seat structures, safety systems, and user scenarios

In other words, the competition is moving from standalone specs to complete systems engineering.

Why This Matters Globally

These three developments are connected more than they first appear.

1. China is stress-testing ADAS in public, comparative ways

Real-world challenge formats like the Beijing route are helping turn smart driving into a measurable competitive category. That creates pressure on brands such as NIO, Xiaomi, Zeekr, Li Auto, and Avatr to prove performance in edge cases, not just product launches.

2. Battery competition is broadening beyond lithium-ion

If CATL’s sodium-ion rollout lands at scale, global automakers will have to rethink entry-EV cost structures, winter-performance assumptions, and chemistry roadmaps. It may also accelerate sodium-ion programs outside China.

3. The cockpit is becoming a strategic profit pool

As EV powertrains commoditize, automakers need new ways to differentiate. Scenario-led cabins, safer zero-gravity seating, and integrated software ecosystems could become as important to consumers as acceleration or range.

The Bigger Picture for Chinese EVs

Put together, this week’s news suggests the Chinese EV market is entering a new phase:

• ADAS is moving from marketing language to comparative real-world testing
• Battery innovation is shifting toward cost resilience, safety, and cold-weather practicality
• Smart cabins are evolving into a central battleground for premium positioning

For overseas observers, the key takeaway is that China’s EV leaders are no longer competing on a single axis. They are building advantage simultaneously in autonomous driving, battery chemistry, supplier ecosystems, and in-cabin experience.

The next step is clear: watch the Beijing smart-driving results closely, monitor whether CATL’s sodium-ion cells actually hit large-scale vehicle programs in 2025, and track how quickly scenario-driven cockpit design moves from luxury flagships into mass-market EVs. That combination—not any one headline spec—may define the next chapter of the global EV market.