Baidu pushes for full-stack AI capabilities as Nvidia chip access expands

Baidu is building out every layer of the AI stack, from custom silicon to frontier models to cloud infrastructure, positioning itself as one of the few companies globally attempting true vertical integration in artificial intelligence.

The push comes at a peculiar moment. After years of US export restrictions that locked Chinese companies out of the most advanced chips, the door has cracked open again. Nvidia H200 GPUs have been approved for export to roughly ten Chinese firms between January and May 2026. Baidu, ByteDance, and Alibaba are among the recipients.

Building the stack from the ground up

At its Create 2026 conference, Baidu laid out a chip roadmap that reads like a declaration of independence from Western semiconductor suppliers. The company’s Kunlunxin division launched the M100 chip in early 2026, with the M300 scheduled for early 2027. Both are designed to handle training and inference workloads.

The financial trajectory here is striking. Analysts project Baidu’s chip sales could increase sixfold to approximately RMB 8 billion, or roughly $1.1 billion, by 2026. Macquarie has pegged the valuation of the Kunlun chip unit at around $28 billion, and Baidu has floated plans for a potential separate listing for the Kunlunxin business.

Baidu has been pursuing a full-stack AI approach since 2011.

Then there’s ERNIE 5.1, Baidu’s latest frontier model released in May 2026. The model uses 94% less pre-training costs than its predecessor and runs on one-third the parameters. It was trained on a Kunlunxin cluster with a 97% training rate.

The DeepSeek effect and China’s new AI playbook

China’s AI landscape shifted meaningfully after DeepSeek released its low-cost frontier model last year. That release briefly sent US tech stocks tumbling and forced a reckoning with the assumption that cutting-edge AI required cutting-edge hardware.

Baidu is taking that lesson further. Rather than just optimizing models to work on constrained hardware, Baidu is simultaneously improving the chips themselves while also gaining selective access to Nvidia’s best hardware. Baidu can use H200s where available while building domestic alternatives that reduce dependence on US export policy decisions.

What this means for investors

A $28 billion valuation for the chip unit, if Macquarie’s estimate holds, would make Kunlunxin a significant semiconductor player in its own right. A separate listing could unlock that value while giving Baidu capital to accelerate development of next-generation silicon.

The ERNIE 5.1 efficiency gains are arguably the more consequential data point for the broader AI industry. A 94% reduction in pre-training costs challenges the prevailing narrative that AI progress requires ever-larger capital expenditures on compute.

The wildcard remains US export policy. The approval of H200 sales to Chinese firms represents a loosening from the strictest periods of the chip embargo, but these decisions are inherently political and subject to reversal. Baidu clearly treats the current access window as potentially temporary, which is precisely why it’s spending billions on chips it could theoretically just buy from Nvidia.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.