Pangu Large Model is a large model platform launched by Huawei Cloud. Its official slogan is "Huawei Pangu Large Model, Reshaping All Industries," emphasizing the application of AI capabilities to various industry scenarios.
I. Positioning and Purpose According to official page information, Pangu Large Model brings together Skills, models, and datasets, and is committed to building an open and efficient AI asset ecosystem. Its target users include developers and enterprise users, and it can be used to build and expand agent applications and provide model and data support for industry scenarios.
II. Main Capabilities
- Skill plugins: The platform provides a rich set of skill plugins to expand unlimited tool boundaries for agents.
- Models and datasets: The platform brings together model and dataset resources. The official page lists datasets such as medical-o1-reasoning-SFT-Datasets.
- Skill installation: Supports installing skills via the npx command. Official examples include: npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-billing-scout npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-find-skills npx skills add https://pangu.huaweicloud.com/skills/ur5217372/dev-design-workflow npx skills add https://pangu.huaweicloud.com/skills/ur5217372/djbh-assessment
- Government application development: The official page mentions the "Government Application Development Requirements Analysis and Design Skill," involving requirements analysis and design workflows.
III. How to Use Users can visit the Pangu Large Model official website (https://pangu.huaweicloud.com/），查阅Skills、模型与数据集资源，并按官方提供的npx命令安装所需技能。关于账号注册、计费方式、API调用方式等细节，官方提供的信息中暂未说明。
IV. Applicable Scenarios Official information shows that it is oriented toward industry applications, covering directions such as government application development, and expands agent capabilities through skill plugins. It is suitable for developers and enterprises that need to build industry agents or use models and datasets.
V. Limitations and Information Not Yet Provided Current materials do not provide specific fee and pricing information, the specific list of models supported by the platform, performance indicators, technical architecture descriptions, or complete usage tutorials. The above content is not yet provided and should be subject to the latest official documentation.
