Zactonz AI Connector for Ollama

Description

This plugin by Zactonz Technologies provides an Ollama connector for the WordPress AI Client. It is built for the WordPress Settings > Connectors screen and for the plugin directory connector search, so site owners can add Ollama when they need a connector for local AI, self-hosted AI, or Ollama Cloud.

Use Zactonz AI Connector for Ollama when WordPress asks you to search the plugin directory for a connector. After activation, Ollama appears as an AI connector and can power text generation, image generation, embeddings, tool calling, structured output, and streaming where supported by the selected model.

It lets WordPress sites use large language models running locally, on a remote self-hosted Ollama instance, or on Ollama Cloud for text and image generation and other AI capabilities.

Disclaimer: Zactonz AI Connector for Ollama is developed by Zactonz Technologies. Ollama is a third-party project. This plugin is not affiliated with, endorsed by, or sponsored by Ollama.

Source code, issues and release history are on GitHub. Setup and usage docs are on the Zactonz developer portal.

Ollama exposes an OpenAI-compatible API, and this connector uses that API to communicate with any model you have pulled into Ollama (Llama, Mistral, Gemma, Phi, and many more).

Features:

• Text generation with any Ollama model
• Image generation with supported models
• Automatic model discovery from your Ollama instance
• Function calling support
• Structured output (JSON mode) support
• Connector integration for Settings > Connectors with an Ollama Cloud or self-hosted connection switch
• Discoverable WordPress AI connector metadata for plugin-directory connector searches
• Separate saved API keys for Ollama Cloud and protected self-hosted endpoints
• Self-hosted host URL and port controls, default model selection, thinking mode, and model discovery from the selected endpoint
• Admin text request timeout for slow local or thinking responses
• Works without an API key for local instances
• Single and batch embedding generation on WordPress 7.1 and newer
• Accurate per-model discovery for text, vision, tools, thinking, images, structured output, and embeddings
• Separate default models for text, vision, images, embeddings, and tool calling
• Default, disabled, low, medium, and high thinking controls
• Redacted diagnostics plus WordPress Site Health integration
• Stable cancellable SSE streaming for content, thinking, and tool-call chunks

Requirements:

• PHP 7.4 or higher
• WordPress 7.0 or higher
• Ollama running locally or on a remote server (like Ollama Cloud)

External services

This plugin connects to Ollama, which may be software you run yourself or the third-party Ollama Cloud service, depending on which connection you choose.

Self-hosted or local Ollama. Requests go to the host and port you configure, http://localhost:11434 by default. Nothing leaves your own machine or network, and no third party is involved.

Ollama Cloud. Choosing the Cloud connection sends requests to the Ollama service at https://ollama.com. What is sent: the API key you configure, and the prompts, images, tool definitions and generation settings that a WordPress AI feature passes to the AI Client, each time such a feature runs a request through this connector. The connector also requests the model catalogue when its settings screen loads, when a WordPress AI feature asks which models are available, and when you run diagnostics or the Site Health test. No other data is sent.

Service provider: Ollama (terms of service, privacy policy).

Technical details