Hologres AI nodes (GPU) provide dedicated compute for model inference, vector search, and intelligent analysis. The latest pricing is available in the Hologres management console.
AI node overview
Hologres AI nodes are GPU-enhanced compute nodes for large model inference, vector search, and intelligent analysis. AI nodes provide you with dedicated GPU resources. With built-in large models, AI Function lets you query data and invoke models without data movement. This reduces transfer costs and supports recommendation, search, intelligent Q&A, and Retrieval-Augmented Generation (RAG) scenarios.
Billing
AI resources are billed by node specification. Two billing methods are available:
-
Subscription: Pay upfront.
-
Pay-as-you-go: Billed hourly. Charges apply even when resources are idle.
Each AI resource includes a 200 GB system disk at no extra cost.
Supported regions
China (Hangzhou); China (Beijing)
Limitations
-
Requires Hologres V3.2 or later. Each instance supports only one AI resource.
-
Only one AI node specification type per purchase. To change specifications, unsubscribe the current AI resource and purchase a new one.
-
Backup and restore does not cover AI resources. Purchase new AI resources for any restored instance.
-
For primary/secondary instances: all AI resource operations (purchase, upgrade/downgrade, unsubscribe, release, model deployment) must be performed on the primary instance. Secondary instances can only view and use the primary instance's AI resources and models.
Purchase an AI resource
On the Hologres Instances page, click Create Instance and configure the following parameters. Purchase a Hologres instance.
|
Parameter |
Description |
|
Product Type |
Select AI Resource (Subscription) or AI Resource (Pay-As-You-Go). |
|
Region |
Select a region where AI resources are available. Note
Zone selection is not required. Hologres handles routing and allocation automatically. |
|
Instance ID Associated with AI Resource |
Select the instance to associate. The AI resource integrates into the instance and activates automatically for AI function calls. |
|
AI Node Specifications |
Select a GPU or PPU (Zhenwu PPU Cloud Service) model for your workload. |
|
AI Node |
Set the number of AI nodes. |
|
Subscription Duration |
Set the subscription duration. Note
Required only for AI Resource (Subscription). |
Pricing
AI resources are priced by node specification under two models: pay-as-you-go and subscription. AI resources.
-
Subscription: Pay upfront.
-
Pay-as-you-go: Billed hourly, regardless of usage.
Each AI resource includes a 200 GB system disk at no extra cost.