Volcengine IGA Pages Adds AI Acceleration Gateway for Model Access

On November 11, 2026, InfoQ China reported that Volcano Engine's IGA Pages had integrated an AI acceleration gateway into its web application hosting capability. IGA Pages sits inside Volcano Engine's site acceleration product family and handles hosting and deployment for web apps. The new integration folds model API key management, server-side proxying, cross-origin handling, and application deployment into a single layer, so that model calls require no proxy development on the developer's side.

According to the same report, connecting a web app to a large model has typically meant building out a server-side call chain first. A model API key cannot sit in frontend code, since it would travel with client-side assets and become exposed. To issue requests from the server, developers must obtain and safeguard the API key, configure the model service address, write forwarding logic, and resolve cross-origin issues between the frontend and the proxy. That proxy layer also has to be deployed and maintained separately, leaving developers to juggle the model service, the server-side runtime, and the hosting platform alongside the page and business logic itself.

With the integration, enabling the AI acceleration gateway at project creation triggers the system to provision a dedicated gateway instance for the account. The model service address and API key are written into the project configuration and can be read directly by application code, so developers no longer apply for keys, stand up proxies, or deal with CORS. Projects support model configuration and environment variable injection, and the gateway speaks the OpenAI protocol, acting as a unified proxy layer between the application and model providers: the app sends model requests to the gateway, which forwards them to the backend model service, so teams do not integrate with each provider's API directly or manage provider-specific differences.

The capability is offered with IGA Pages at no charge to start: each account gets a 500K token free quota, and Pages dedicated gateway instances incur no gateway traffic or request fees. The report covers the feature and quota terms only; details on launch timing, supported models, and what comes next are not provided.