Skip to main content
Providers

Inference

Use inference/<model> with VoltAgent's model router.

Quick start​

import { Agent } from "@voltagent/core";

const agent = new Agent({
  name: "inference-agent",
  instructions: "You are a helpful assistant",
  model: "inference/google/gemma-3",
});

Environment variables​

  • INFERENCE_API_KEY

Provider package​

@ai-sdk/openai-compatible

This provider uses the OpenAI-compatible adapter.

Default base URL​

https://inference.net/v1

You can override the base URL by setting INFERENCE_BASE_URL.

Provider docs​

Models​

Show models (9)
  • google/gemma-3
  • meta/llama-3.1-8b-instruct
  • meta/llama-3.2-11b-vision-instruct
  • meta/llama-3.2-1b-instruct
  • meta/llama-3.2-3b-instruct
  • mistral/mistral-nemo-12b-instruct
  • osmosis/osmosis-structure-0.6b
  • qwen/qwen-2.5-7b-vision-instruct
  • qwen/qwen3-embedding-4b

Table of Contents