Skip to content

Model providers

Genkit talks to models through provider plugins. You configure a plugin once, then call any model it exposes through the same generate API. Because the interface is the same across providers, you can swap one model for another, or combine several in one app, without rewriting your application code.

Available providers

Section titled “Available providers”

Between them, these plugins reach Gemini, Claude, GPT, Grok, DeepSeek, Qwen, Kimi, GLM, Llama, and Mistral, plus local models through Ollama, anything routed through OpenRouter, and any other service that exposes an OpenAI-compatible endpoint.

  • Google Generative AI: Gemini models through the Google AI Studio API. This is the fastest way to start, using a single API key.
  • Google Vertex AI: Gemini and other models through Google Cloud, with IAM-based auth for production workloads or an API key for Express Mode.
  • Vertex AI Model Garden: Claude, Llama, and Mistral models hosted on Vertex AI, one plugin per family.
  • Anthropic (Claude): Claude models through the Anthropic Messages API.
  • OpenAI: GPT models and embedders through the OpenAI API.
  • Azure AI Foundry: Models hosted on Azure.
  • AWS Bedrock: Models hosted on AWS.
  • xAI (Grok): Grok models through the xAI API.
  • DeepSeek: DeepSeek models.
  • OpenRouter: One gateway to models from many vendors, with provider routing and fallback chains.
  • Kimi: Moonshot AI’s Kimi models.
  • z.AI: Z.ai’s GLM text and vision models.
  • DashScope (Qwen): Alibaba Cloud’s Qwen models.

Local models

Section titled “Local models”
  • Ollama: Run open models such as Gemma and Llama locally, with no API key.

OpenAI-compatible APIs

Section titled “OpenAI-compatible APIs”