Skip to main content

AI

The Arvancloud AI service (AIaaS) helps developers and teams access powerful language models (LLMs) through secure, customizable AI gateways. With this service, you can choose the models you need from the available LLMs in a gateway and send requests to them with dedicated access keys. The AI Gateway guide helps you get started with this service.

Arvancloud AIaaS is designed with a focus on flexibility, ease of use, and scalability, and reduces the infrastructure complexity of using LLMs in production environments.

Features​

The most important features of the Arvancloud AI service include:

  • AI gateway: create a shared access point for multiple models, with a free Arvancloud domain or your own domain.
  • Model selection: easily choose from a list of high-performance LLMs through a web interface.
  • Secure access: access the API only with access keys generated by the user.
  • Token usage monitoring: see text reports and interactive charts of input/output token usage and model costs.
  • Request rate limit: set a cap on the number of requests in specific time periods for each gateway.
  • Hosted models: some models are hosted directly on Arvancloud infrastructure, so you get optimal performance and lower latency.
  • Knowledge base: by taking in your own data, this feature lets you get AI answers based on your own data.
  • Guardrail: protect requests and responses against sensitive data leaks, jailbreaks, and inappropriate content.

Limitations​

  • Fine-tuning or training models isn't currently available.
  • Conversation history and stateful interactions between API requests aren't stored.

Note that the Arvancloud AI service doesn't store any data from users' conversation history.