Huawei's AI Agent Xiaoyi Work First Uses openPangu-2.1-Pro Model to Enhance HarmonyOS Tasks

DeepBlue Technology launched Wowoker on October 29th—a corporate work phone developed in collaboration with Huawei. The Chinese company, specializing in artificial intelligence, positions this gadget as the world's first work phone featuring a large agent model.

According to DeepBlue, the device operates on Huawei Kirin chips, utilizes Huawei hardware, and runs on the HarmonyOS operating system, which is integrated with DeepBlue's multimodal large model, its Decision Making Model, and the AgentOS orchestration system.

Huawei provided the HarmonyOS ecosystem, the hardware adaptation basis, and the foundations of corporate security, offering 12 phone models for customization. This allows for selecting suitable equipment for teams ranging from ten people up to groups purchasing hundreds of devices.

Device Focus on Sales Operations

The phone is exclusively oriented towards sales teams, not end consumers. Built-in agents perform tasks such as lead generation, automated outbound calls, and buyer intent analysis. The lead generation agent accepts industry settings and keywords from the manager, then searches for potential clients on Xiaohongshu, WeChat, and Douyin, initiating dialogue with them.

DeepBlue reported that during testing with an early client, the OPC community of Greenland, the agent was able to process over 200 potential leads in half a day, more than 50 of which led to substantive conversations. The presentation demonstrated the operation of an AI agent that called a parent to confirm the child's age, offered a trial lesson, and organized subsequent communication via WeChat.

Additional features include the ability to make and receive calls using a voice mimicking the owner's voice, publishing content across various platforms, tagging clients, and screening leads. On the management side, all calls, text messages, and chats in WeChat, WeCom, Feishu, and Douyin are encrypted and synchronized with the company's cloud. Furthermore, the AI conducts quality checks, noting incorrect phrasing in real time.

Client contact details and WeChat IDs are stored only on the company's backend, so salespeople see masked details, cannot delete chat history, and cannot save contacts personally. Upon an employee's departure, the devices can be remotely disabled.

Reasons for Choosing Hardware Solution

When asked about the reason for creating dedicated hardware instead of an application, DeepBlue explained that a significant portion of sales department work occurs within third-party applications where cloud software is limited by system permissions and interfaces. The phone, however, provides vendor-level adaptation and retains some data on the device itself for processing. Chairman Chen Haibo noted that business clients with real needs are more likely to receive support from major platforms.

Regarding concerns about intrusive calls, the company stated that the system has built-in rules that block high-frequency outbound calls, terminate contact if the client explicitly refuses, and allow the user to take control at any moment.

Wowoker is not sold individually; companies purchase a three-year agent service package that includes the phone. The cost of this package is 5,299 RMB or 4,999 RMB initially, with no additional fees throughout the term.

Huawei Cloud has announced plans to decommission the openPangu-2.0-Flash and openPangu-2.0-Pro model services on its MaaS (Model as a Service) platform. According to the model deprecation notice, these services will stop being sold on November 6th Beijing time and will be completely disabled on November 30th.

IT Home reported this notice, which was published on the Huawei Cloud support website on November 3rd. As part of the schedule, new users will be unable to purchase or activate the specified services after November 6th, but existing customers will retain the ability to use them until the end of the month. After November 30th, access to both models will be closed for both new and existing users. The notice lists these services in the CN Southwest-Guiyang1 Huawei Cloud region.

The company offers recommended replacements for each model. Users of openPangu-2.0-Flash version 20260630 are advised to switch to DeepSeek-V4.1-Flash-1017, while users of openPangu-2.0-Pro version 20260730 are directed to Zhipu's GLM-5.3. Huawei Cloud emphasizes that the models differ in capabilities, API parameters, pricing, and real-world performance, and advises customers to review the documentation and conduct their own assessment before switching.

The migration guide indicates that customers using the standard MaaS V1 API will need to switch to an OpenAI-compatible endpoint, whereas those already using API V2 or an OpenAI-compatible interface only need to change the model parameter.

According to IT Home, Huawei Cloud plans to release a new generation of openPangu models incrementally but has not yet disclosed their names or timelines. These two models were among Huawei's largest open releases this year: openPangu-2.0-Flash with 92 billion parameters was released at the end of June, and the sparse openPangu-2.0-Pro model with 505 billion parameters, trained on Ascend NPUs, was released at the end of July. In September, Huawei also released code for pre-training, controlled fine-tuning, and reinforcement learning for this series. However, the deprecation notice applies only to hosted MaaS services and does not affect open releases.

Model changes on the platform happen frequently. In the same bulletin, Huawei Cloud scheduled the decommissioning of GLM-5.1 on November 16th, replacing it with GLM-5.3, and is also migrating users from DeepSeek-V4-Flash to DeepSeek-V4.1-Flash and decommissioning Kimi-K2.6 on November 8th, while also listing several Qwen and Wan models for retirement.

Previously, DeepSeek-V4-Pro was decommissioned in September, and several versions of DeepSeek, Qwen, and GLM were retired in the summer. For MaaS customers, this means that workloads built on openPangu-2.0 must be migrated to other hosted platform models, at least until the next generation of openPangu is available.

Xiaomi has announced the release of the Xiaomi-OCR-0 model, a lightweight document understanding model. The weights and inference code for this model are available on Hugging Face and GitHub, the company reported on October 30th.

The model boasts 0.8 billion parameters and is designed to perform three types of tasks within a single system: document parsing, key information extraction, and visual question answering via OCR.

The parsing functionality includes recognizing main text, tables, and mathematical formulas, outputting structured results in reading order with support for native Markdown and LaTeX formats. For key information extraction, the model focuses on invoices, receipts, IDs, and business forms, locating requested fields and providing the result in JSON format for automated workflows.

Visual question answering goes beyond simple transcription, enabling cross-regional searching, numerical calculations, and reasoning based on documents and charts.

Training and Methodology

Xiaomi states that the model is built upon the Qwen3.5-0.8B-Base model and was trained on approximately 170 million OCR and document samples. Data labels were filtered by achieving consensus among several expert OCR models that performed mutual verification and voted on areas of text, tables, and formulas.

When working with complex tables and formulas, the team utilized a 'predict, re-render, and compare' loop: candidate results were re-rendered into images and compared with the original at the pixel and structural level to detect shifted rows or missing characters. Synthetic complex cases included multilingual text, historical handwriting, screen moiré effects, and distorted lighting.

The training process occurred in three stages: first, textual fine-tuning to align visual regions with text tokens; second, continued pre-training of OCR on tens of billions of tokens; and finally, reinforcement learning for mixed tasks with verifiable rewards, such as edit distance and table matching scores.

Testing Results

On the OmniDocBench v1.6 benchmark, the model achieved a score of 96.83, ranking second overall, only 0.04 points behind the leader, and becoming the top performer among models with less than one billion parameters. In the Real5-OmniDocBench test, which simulates scanning, crumpled paper, screenshots, lighting variations, and tilted frames, the model scored 95.24. Its performance on Wild-OmniDocBench was 87.94.

The model reaches 97.84 on the UniMER-Test reference formula test and 92.27 on PubTabNet for table structure evaluation. Averaging results across DocVQA, InfoVQA, ChartQA, OCRBench, and TextVQA, it scores 83.2, surpassing HunyuanOCR-1.5 with 76.5 and 8B MiniCPM-V-4.5 with 82.6, although it trails Qwen3.5-4B with a score of 84.9. It should be noted that all presented figures are proprietary data from Xiaomi.

Xiaomi reports that this technology is already integrated into the Super XiaoAi assistant on its phones, where it can extract boarding passes, queue numbers, bill details, and meeting times directly from the screen. In addition to the model weights, Xiaomi provides developers with a browser demo version, the MCP service, and an Agent Skill toolkit.