Gongji Cloud AI Inference Serverless Platform (access address:https://console.suanli.cn/auth/login）是一个面向AI推理任务的Serverless GPU computing platform.
Main purpose This platform is aimed at users who need to run AI inference workloads, such as deploying and calling machine learning or deep learning models for online inference. Its Serverless form means users do not need to prepare and manage GPU servers themselves, and can use computing power on demand. It is suitable for scenarios where inference task volume fluctuates greatly and users want to reduce operations and maintenance investment.
Main capabilities According to existing information, the platform's core capabilities include:
- Extreme elasticity: elastically scale computing resources on demand in a Serverless manner to adapt to changes in inference request volume.
- Minimal deployment: simplify the deployment process for GPU environments and inference services, lowering the barrier to entry.
- Abundant resources: provide a variety of GPU resource options.
- Low price: provide RTX 4090 computing power, with prices as low as 1.68 yuan/hour.
How to use Users can log in and manage related resources through the console address https://console.suanli.cn/auth/login . Regarding specific operation steps such as the registration process, creating inference services, configuring models, and calling interfaces, the information is not currently provided. Please refer to the actual page instructions in the platform console.
Applicable scenarios Suitable for developers, teams, or enterprises that need to run AI inference tasks on demand at lower cost, especially scenarios where inference workloads have obvious peaks and troughs, where users want to avoid the cost of long-term idle GPUs, and where users are unwilling to invest a large amount of effort in maintaining GPU infrastructure.
Limitations and information not currently provided The current information does not explain the model frameworks supported by the platform, image environments, storage solutions, concurrency and quota limits, available regions, billing details (such as whether billing is per second and whether there is a minimum charge), free trial credits, SLA guarantees, etc. The above content is not currently provided. Before use, it is recommended to consult the official documentation or console instructions.
