Chinese smartphone makers are rebuilding their operating systems around AI agents that can plan and execute tasks across apps, reports Caixin. The manufacturers are placing a new bet on software as falling shipments and rising prices weigh on the global handset market.

Oppo, Vivo and Honor Device unveiled new operating systems at separate developer conferences last week, each promising to turn the smartphone from a device that responds to commands into an assistant capable of anticipating users’ needs and carrying out multistep tasks.

The announcements mark the latest phase of the mobile industry’s AI push. Phone makers initially focused on stand-alone chatbots and features such as photo editing and text generation. They are now integrating AI deeper into their operating systems, giving agents access to apps, personal data and device functions.

Ant International, the overseas affiliate of Chinese fintech giant Ant Group, is making a major bid to power the next phase of mobile payments: letting autonomous AI agents handle your electronic wallet to make purchases, reports the South China Morning Post.

The company on Friday open-sourced its Agentic Mobile Protocol (AMP) on GitHub, effectively putting it forward as a potential global technical standard for how AI agents securely pay for real-world goods. AMP was built onto Ant International’s existing Alipay+ cross-border network, which connects to 10 major Asian digital wallets serving 1.5 billion consumer accounts for an initial phase, according to the company.

The protocol’s release would “enable agents to enter existing mobile payment systems as trusted actors”, allowing platforms like Alipay in mainland China, AlipayHK in Hong Kong, Starryblu in Singapore, and KakaoPay or Toss in South Korea to accelerate their AI strategies, said Yang Jiang-ming, Ant International’s chief innovation officer.

DeepSeek released a new smaller model on Thursday that it says can outperform some flagship systems while cutting inference costs, reports Caixin. This comes as Chinese AI developers increasingly compete on both model capability and price.

The new DeepSeek-V4.1-Flash is a 552-billion-parameter mixture-of-experts model with native multimodal visual understanding. DeepSeek said the model uses a new architecture designed to deliver higher performance, faster inference, higher throughput and easier scaling to larger models.

DeepSeek said V4.1-Flash outperformed domestic rivals including Moonshot AI’s Kimi K3 and Z.AI’s GLM-5.3 on benchmarks such as Terminal-Bench 3.0, DeepSWE v1.1, CyberGym and Automation-Bench. It also beat Anthropic’s Opus 5 and OpenAI’s GPT-5.6 Sol on some tests, according to the company.

Chinese AI chipmakers including Huawei and Cambricon have sharply raised prices for current and next-generation AI processors, reports Reuters citing inside sources. This comes as soaring costs for high-bandwidth memory squeeze efforts to build domestic alternatives to Nvidia.

Huawei has lifted the indicated price of its Ascend 950DT accelerator card, which integrates an AI processor with memory and other components, to above RMB 250,000 ($37,255), two people said. The increase represents a jump of 20% to 50% from prices quoted to customers just two months ​ago, depending on contract terms, they said. 

Beijing-based Cambricon has also repriced its next-generation chip, tentatively called the 690, at 20% to 30% ⁠above levels it indicated two months ago, two sources said, without giving an exact figure. Smaller rivals MetaX and Iluvatar CoreX have made similar increases, the ​sources said.

China’s Moonshot AI is ​negotiating revenue-sharing agreements with Microsoft , Amazon and Alphabet’s Google that would allow the US cloud giants to host its ‌blockbuster Kimi K3 model, reports Reuters citing inside sources. Any such deal could mark the first big revenue-sharing pact between a Chinese AI firm and a major US cloud company.

IPO-bound Moonshot is seeking up to a 30% share of revenue generated from K3-related services ⁠on Microsoft’s Azure, Amazon Web Services and Google Cloud, according to the sources who declined to be identified because the discussions are private. That ​would be in line with terms that sources have said the startup has outlined for major customers using the open-weight model. The discussions are at ​an early stage and there is no certainty they will result in agreements, the sources said.

The discussions highlight how China’s leading AI models, often far cheaper than Western offerings, are gaining traction in the US, despite national security concerns in Washington that ​have led to bans on exports of AI chips to China. They are also taking place despite critical comments about Moonshot ​from senior US officials.