Skip to content
AI Puffer manual

Settings

Everything on the right saves into the template you’re using, so a good setup is there next time.

Settings

Section titled “Settings”
  • Template: start from Short, Medium or Long, or your own. New template saves your setup; Restore starter templates brings the originals back. A starter template changes only the prompts and length, not your model or publishing choices.
  • Model: AI Puffer Cloud on credits, or your own key for OpenAI, Google, Anthropic, OpenRouter, Azure, DeepSeek or xAI. Ollama is Pro. The gear has Temperature and, for models that think, Reasoning.
  • Length: short, medium or long. It sets how long the answer may be, not a word count; ask for words in the content prompt.
  • Status: draft, publish, pending review or private.

Prompts

Section titled “Prompts”

One row per part: Title, Content, Meta description, Focus keyword, Excerpt and Tags. Switch parts on or off, and open Instructions to change how each is written. Choose a preset, keep your own in the Library, and use variables such as {topic} and {keywords}. Each part is a separate AI request.

SEO

Section titled “SEO”
  • Smart SEO Pro: checks the article against your SEO plugin (Yoast SEO, Rank Math or All in One SEO) and rewrites it until it scores well. Choose an approach: Strict score, Balanced or Content first.
  • Table of contents: adds links to the headings at the top.
  • Optimize URL: a short slug from the focus keyword.

The SEO description and focus keyword are saved into your SEO plugin’s own fields.

Images

Section titled “Images”
  • Content images and Featured image, each with its own instructions.
  • Image source: AI images from OpenAI, AI Puffer Cloud, Google, OpenRouter, Azure, xAI or Replicate, or stock photos from Pexels or Pixabay. Keys for Replicate, Pexels and Pixabay go under Settings › Connections.
  • Images per article (up to 10) and Placement, such as after the first H2 or every few paragraphs.
  • Image options: size, alignment, and the provider’s own choices.
  • The images’ title, alt text, caption and description can be written too.

Context

Section titled “Context”

Let the writer draw on your own knowledge: choose a store under Context (built-in, OpenAI, Google, Pinecone, Qdrant or Chroma) and how many pieces it reads. See the chatbot’s knowledge page for what each store is.

Publishing

Section titled “Publishing”

Content format (Gutenberg blocks or Classic), Post type, Author and Categories.