Hoosh Logo
Hoosh
Documentation

Hoosh AI Settings Guide

Everything you need to configure and customize your AI writing assistant for WordPress.

Getting Started

Quick setup guide to get Hoosh running on your WordPress site

Installation

  1. 1Upload the hoosh-ai-assistant folder to /wp-content/plugins/
  2. 2Activate the plugin through the Plugins menu in WordPress
  3. 3Navigate to Settings → Hoosh AI and enter your API key
  4. 4Start creating content with AI assistance!

Requirements

  • WordPress 6.0 or higher
  • PHP 8.2 or higher
  • Gutenberg (Block) editor
  • At least one API key (OpenAI or Google Gemini)

API Configuration

Connect Hoosh to OpenAI or Google Gemini

Choose Your Provider

Hoosh supports multiple AI providers. Select your preferred provider in settings and enter the corresponding API key.

🤖OpenAI

GPT-4o, GPT-3.5 Turbo for text • DALL-E 3, DALL-E 2 for images

✨Google Gemini

Gemini 2.5 Flash Lite, Gemini 3 Flash Preview for text • Nano Banana for images

OpenAI API Key

Your OpenAI API key for accessing GPT models and DALL-E image generation.

How to get your OpenAI API key:

  1. 1. Visit platform.openai.com/api-keys
  2. 2. Sign in or create an OpenAI account
  3. 3. Click "Create new secret key"
  4. 4. Copy and paste the key into Hoosh settings

Google Gemini API Key

Your Google Gemini API key for accessing Gemini models and Nano Banana image generation.

How to get your Gemini API key:

  1. 1. Visit aistudio.google.com/apikey
  2. 2. Sign in with your Google account
  3. 3. Click "Create API key"
  4. 4. Copy and paste the key into Hoosh settings

🔒Your API keys are stored securely in the WordPress database and are never exposed in the browser. You can switch between providers at any time without losing your keys.

Model Selection

Choose the AI models for text and image generation

Available Models by Provider

The available models depend on your selected AI provider. Each provider offers different text and image generation capabilities.

Text Models

Select the model used for text generation tasks like title completion, content suggestions, and excerpts.

OpenAI Models

GPT-4o

Latest model with best quality and reasoning capabilities.

GPT-3.5 Turbo

Fast and cost-effective. Good for simple completions and high-volume usage.

Google Gemini Models

Gemini 2.5 Flash Lite

Fast and efficient model optimized for quick responses. Great balance of speed and quality.

Gemini 3 Flash Preview

Latest preview model with enhanced capabilities. May have occasional changes.

Image Models

Select the model for generating featured images and thumbnails.

OpenAI Models (DALL-E)

DALL-E 3

Higher quality images with better prompt understanding. Generates 1024×1024 images.

DALL-E 2

Faster generation at lower cost. Good for simpler image needs.

Google Gemini Models (Nano Banana)

Nano Banana

Fast image generation powered by Google Gemini. Good for quick thumbnails.

Nano Banana Pro

Enhanced image quality with better detail and composition.

Usage Limits

Control API costs with configurable limits

Set usage limits to control your API spending. When a limit is reached, AI generation will be paused until the limit resets. Set any limit to 0 for unlimited usage.

Daily Token Limit

Maximum number of tokens that can be used per day across all AI operations.

Default: 100,000 tokens

💡Tokens are the basic units AI models use to process text. A typical blog post might use 1,000-3,000 tokens for completion suggestions.

Monthly Token Limit

Maximum number of tokens allowed per month. Resets on the first day of each month.

Default: 1,000,000 tokens

Daily Request Limit

Maximum number of AI requests (API calls) per day, regardless of token usage.

Default: 50 requests

Monthly Request Limit

Maximum number of AI requests allowed per month.

Default: 1,000 requests

Thumbnail Generation Limit

Maximum number of AI-generated images per month. Image generation typically costs more than text generation.

Default: 0 (unlimited)

Features

Enable, disable, and customize AI-powered features

Title Auto-Completion

Toggle

Get real-time title suggestions as you type in the post title field.

Default: Enabled

Title Generation Prompt

Template used when generating title suggestions.

Write a catchy title for a post about: {content}
Available variables:{content}

Content Auto-Completion

Toggle

Press Tab in a paragraph block to get AI-powered content suggestions that continue your thought.

Default: Enabled

Content Suggestions Prompt

Template for generating content continuations.

Suggest content ideas for a post titled: {title}
Available variables:{title}{content}

Thumbnail Generation

Toggle

Adds a 'Generate with AI' button to the Featured Image panel in the editor.

Default: Disabled

Thumbnail Prompt

Template for generating featured images.

Generate a featured image for: {title}
Available variables:{title}{content}

Summary Generation

Toggle

Adds a 'Generate Excerpt' button to the Excerpt panel for one-click summary creation.

Default: Disabled

Summary Prompt

Template for generating post excerpts.

Summarize the following content in 2-3 sentences: {content}
Available variables:{content}

Tag Suggestions

Toggle

Adds an 'Add 5 Tags' button to the Tags panel for AI-powered tag recommendations.

Default: Enabled

Tag Suggestion Prompt

Template for generating tag suggestions.

Generate {count} relevant tags for the following post. Prioritize existing tags when applicable.
Available variables:{count}{title}{content}{existing_tags}

Show 'Press Tab' Hint

Toggle

Display a subtle hint next to AI suggestions reminding users to press Tab to accept.

Default: Enabled

Learning Engine

AI that learns from your content to provide better suggestions

The Learning Engine analyzes your published posts to understand your blog's themes, topics, and writing style. This context is used to improve taxonomy suggestions and ensure AI recommendations match your content strategy.

Context Memory

The AI-generated summary of your site's content style, tone, and topics. This context is included in prompts to personalize suggestions.

What's stored in Context Memory:

  • • Main topics and themes of your blog
  • • Writing style and tone analysis
  • • Common categories and tag patterns
  • • Content structure preferences

✨You can manually edit the context memory to refine the AI's understanding of your site's voice and focus.

Regenerate Context

Re-analyze all published posts to update the context memory. Useful when your content focus has shifted or after publishing significant new content.

Note: Regenerating context may take a few moments depending on the number of published posts. The process uses your API tokens.