README

A WordPress plugin that provides local and cloud AI model support (Ollama) for the WordPress AI Client.

Description

WP Ollama Model Provider enables WordPress to use AI models through Ollama, supporting both local installations and Ollama Cloud. This plugin acts as a provider for the WordPress AI Client, making Ollama models accessible to any WordPress plugin that uses the AI Client.

Features

• Flexible Deployment: Support for both local Ollama installations and Ollama Cloud
• Local AI Models: Run AI models locally with Ollama - no cloud API keys required
• Cloud Integration: Use Ollama Cloud for easy access without local installation
• Automatic Model Detection: Discovers all available Ollama models on your system or cloud account
• Simple Configuration: Easy settings page at Settings > Ollama AI Models
• Model Selection: Choose which Ollama model to use from a dropdown
• Model Caching: Efficient 5-minute cache for model discovery
• Public API: Other plugins can easily check for and use your selected model
• Privacy-First: Local mode ensures text generation happens entirely on your machine

Supported Providers

• Ollama (current)
  • Local installation
  • Ollama Cloud
• Future: LocalAI, LM Studio, and other local providers

Requirements

• PHP: 8.0 or higher
• WordPress: 6.0 or higher
• Dependencies:
  • wordpress/wp-ai-client ^0.2.1
  • For local mode: Ollama installed and running locally
  • For cloud mode: Ollama Cloud API key

Installation

Via Composer (Recommended)

# Install in your WordPress project
composer require jonathanbossenger/wp-ollama-model-provider

The plugin will be installed to wp-content/plugins/wp-ollama-model-provider/ (or web/app/plugins/ for Bedrock). Activate it through the WordPress admin.

For Bedrock/Roots.io users: The package will automatically install to the correct plugins directory.

For traditional WordPress installations: Run composer from your WordPress root with proper installer-paths configured, or run it directly in wp-content/plugins/.

Manual Installation

1. Download the latest release from GitHub Releases
2. Upload to /wp-content/plugins/wp-ollama-model-provider/
3. Run composer install --no-dev in the plugin directory
4. Activate the plugin through the WordPress admin

Setup Ollama

Local Setup

1. Install Ollama from https://ollama.com
2. Pull at least one model:

   ollama pull llama3.2

3. Verify Ollama is running:

   curl http://localhost:11434/api/tags

Ollama Cloud Setup

1. Sign up for Ollama Cloud at https://ollama.com/cloud
2. Get your API key from the Ollama Cloud dashboard

Configuration

For Local Ollama:

1. Navigate to Settings > Ollama AI Models in WordPress admin
2. Select Local as the deployment mode
3. Select your preferred Ollama model from the dropdown
4. Click Save Settings

For Ollama Cloud:

1. Navigate to Settings > Ollama AI Models in WordPress admin
2. Select Ollama Cloud as the deployment mode
3. Enter your Ollama Cloud API key
4. Select your preferred model from the dropdown
5. Click Save Settings

Your selected model is now available to all WordPress plugins that use the AI Client.

Usage

For End Users

Once configured, the plugin runs automatically in the background. Any WordPress plugin that uses the WordPress AI Client can detect and use your selected Ollama model.

For Plugin Developers

Check for Selected Model

if ( function_exists( 'wp_ollama_model_provider_get_selected_model' ) ) {
    $selected_model = wp_ollama_model_provider_get_selected_model( 'ollama' );

    if ( ! empty( $selected_model ) ) {
        // Use the selected Ollama model
    }
}

Use the Selected Model

// Check for selected Ollama model from wp-ollama-model-provider
if ( function_exists( 'wp_ollama_model_provider_get_selected_model' ) ) {
    $selected_ollama_model = wp_ollama_model_provider_get_selected_model( 'ollama' );

    if ( ! empty( $selected_ollama_model ) &&
         class_exists( 'WpOllamaModelProvider\Providers\Ollama\OllamaProvider' ) ) {
        $model = \WpOllamaModelProvider\Providers\Ollama\OllamaProvider::model( $selected_ollama_model );
        $registry = \WordPress\AiClient\AiClient::defaultRegistry();
        $registry->bindModelDependencies( $model );

        return \WordPress\AI_Client\AI_Client::prompt( $prompt )
            ->using_model( $model )
            ->generate_text();
    }
}

// Fall back to automatic provider/model selection
return \WordPress\AI_Client\AI_Client::prompt( $prompt )->generate_text();

Public API Functions

wp_ollama_model_provider_is_provider_registered( $provider_slug )

• Check if a provider (e.g., 'ollama') is registered
• Returns: bool

wp_ollama_model_provider_has_settings_page()

• Check if the settings page is available
• Returns: bool

wp_ollama_model_provider_get_selected_model( $provider_slug )

• Get the selected model for a provider (e.g., 'ollama')
• Returns: string (model ID) or empty string if none selected

Filters

wp_ai_client_ollama_base_url

Change the Ollama base URL (default: http://localhost:11434).

add_filter( 'wp_ai_client_ollama_base_url', function() {
    return 'http://192.168.1.100:11434'; // Remote Ollama server
} );

wp_ai_client_default_request_timeout

Adjust timeout for slower models (default: varies by plugin).

add_filter( 'wp_ai_client_default_request_timeout', function() {
    return 120; // 2 minutes
} );

Documentation

• Quick Start Guide - Get up and running in 5 minutes
• Ollama Models Reference - Model recommendations and comparisons
• Integration Guide - Technical integration details

Troubleshooting

No models appearing in dropdown

1. Verify Ollama is running:

   ollama list

2. Pull a model if needed:

   ollama pull llama3.2

3. Use the "Refresh Model List" button on the settings page

Cannot connect to Ollama

1. Check if Ollama is running:

   curl http://localhost:11434/api/tags

2. Start Ollama if needed:

   ollama serve

Slow generation times

• Use a smaller model (e.g., llama3.2:1b instead of llama2:13b)
• Increase the timeout using the wp_ai_client_default_request_timeout filter
• See OLLAMA-MODELS.md for model recommendations

Enable debug logging

// In wp-config.php
define( 'WP_DEBUG', true );
define( 'WP_DEBUG_LOG', true );
define( 'WP_DEBUG_DISPLAY', false );

Check wp-content/debug.log for error messages.

Limitations

• Text Generation Only: Ollama only supports text generation, not image generation
• Local Models: Requires Ollama to be installed and running locally (or on your network)
• Resource Intensive: Larger models require significant RAM and CPU/GPU resources

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

Development Setup

1. Clone the repository
2. Run composer install
3. Ensure Ollama is installed and running
4. Activate the plugin in a WordPress development environment

Coding Standards

This plugin follows WordPress Coding Standards. Check your code:

vendor/bin/phpcs --standard=WordPress wp-ollama-model-provider.php includes/

License

GPL-2.0-or-later

Support

• Issues: GitHub Issues
• Documentation: See docs/ directory
• Ollama: https://ollama.com

Credits

Created by Jonathan Bossenger

Built on top of:

• WordPress AI Client
• Ollama