Description
Kwik AI brings AI-powered content automation to your WordPress site. Generate tags, write descriptions, craft excerpts, and create featured images — all from within the WordPress editor. It supports multiple AI providers: Ollama (local models for full data privacy), OpenRouter (hundreds of cloud models via one API), OpenAI (GPT models), and FAL.AI (AI image generation).
Features
- AI Tags Generation — Generate relevant tags from both image and text content, with a preview-and-edit step before applying. Reads images from attachments, Gutenberg blocks, and galleries.
- AI Description from Images (Meta Box) — Analyze post images with vision models and generate engaging descriptions of the right length. Supports configurable word counts and multiple image formats.
- AI Description from URLs (Meta Box) — Scrape and summarize web pages, then generate descriptions based on that content. Useful when your post links to external resources.
- AI Excerpt Generation — Generate post excerpts from the title and content in four tones: technical, straight, excerpt-style, or pithy. Length is adjustable from 20 to 80 words.
- AI Description Block — A native Gutenberg block with built-in generation controls. Choose from image-based, URL-based, or content-based generation. Customize length (50-500 words) and tone. Content stays inside the block.
- AI Featured Image Generation (FAL.AI) — Two-step workflow: first the AI distills your post into a visual theme, then crafts a creative image prompt. Preview and edit the prompt before submitting. Choose from FLUX, Stable Diffusion, Ideogram, Recraft, and more. Configure image size (square, portrait, landscape).
- Multiple AI Providers — Ollama (run models locally), OpenRouter (hundreds of models via one API), OpenAI (GPT models), and FAL.AI (image generation). Use vision models for image analysis and text models for prompts and descriptions.
- Dedicated Text Model — Set a separate text model for tasks like prompt crafting and excerpts, independent from the main vision model. Useful when your primary model is vision-only.
The plugin connects to third-party AI services you configure. Tags, descriptions, excerpts, and images are generated by sending your post content (text, images, or both) to the provider you select.
Third-party services
This plugin can send your post content to whichever provider you configure:
- Ollama — self-hosted; data goes to the host you configure. https://ollama.com/
- OpenRouter — https://openrouter.ai/ (Terms: https://openrouter.ai/terms , Privacy: https://openrouter.ai/privacy)
- OpenAI — https://openai.com/ (Terms: https://openai.com/policies/terms-of-use/ , Privacy: https://openai.com/policies/privacy-policy/)
- FAL.AI (optional, for featured-image generation) — https://fal.ai/ (Terms: https://fal.ai/terms , Privacy: https://fal.ai/privacy)
Screenshots
Blocks
This plugin provides 1 block.
- Description
Installation
- Upload the plugin to your WordPress site and activate it.
- Go to Settings > Kwik AI.
- Select your AI provider (Ollama, OpenRouter, or OpenAI), set the endpoint/model, and enter any API credentials.
- For image analysis, use a vision-capable model (e.g.,
gemma3:27borllava:13bfor Ollama;gpt-4ofor OpenAI). - (Optional) Add a FAL.AI API key and choose an image model to enable AI featured images.
- (Optional) Set a dedicated Text Model for text-only tasks like prompt crafting and excerpts.
- Enable the post types you want Kwik AI to appear on.
FAQ
-
Do I need an API key?
-
For OpenRouter and OpenAI you need an API key. For Ollama or local LLM you only need a running Ollama server (local or remote). FAL.AI requires its own key and is only needed for featured-image generation.
-
Does this send my content to a third party?
-
Yes, when you generate tags, descriptions, excerpts, or images, the relevant post content is sent to the AI provider you configure. See the third-party services list in the Description for links to each provider’s terms and privacy policy.
-
What models work for image analysis?
-
Any vision-capable model: for example
gemma3:27borllava:13bon Ollama,openai/gpt-4oon OpenRouter, orgpt-4o/gpt-4o-minion OpenAI. -
Can I use a different model for text-only tasks?
-
Yes. In settings you can set a dedicated Text Model for text-only generation (image prompts, excerpts, URL-based descriptions) when your main model is vision-only.
-
Which image models are available for featured images?
-
FAL.AI offers many text-to-image models. The plugin includes popular choices out of the box: FLUX.1 (schnell, dev, pro), Stable Diffusion 3.5, Ideogram V2, and Recraft V3. The list is populated from FAL.AI’s live catalog and cached for 12 hours, so new models appear automatically. You can also enter a custom model ID.
-
How does featured image generation work?
-
It uses a two-step process. First, the AI reads your entire post and distills it into a visual theme. Second, that theme is turned into a creative image prompt. You can preview and edit the prompt before submitting it to FAL.AI. The image is generated asynchronously via FAL.AI’s queue API, then sideloaded into your media library and set as the post’s featured image.
-
What image sizes can I generate?
-
You can choose from six preset sizes: Square HD (1:1), Square (1:1), Portrait 4:3, Portrait 16:9, Landscape 4:3, and Landscape 16:9. The default is Landscape 16:9.
-
Can I generate descriptions from URLs?
-
Yes. Paste one or more URLs into the description meta box and Kwik AI will scrape the page content, summarize it, and generate a description based on that summary.
-
What excerpt tones are available?
-
Four tones: Technical (precise and factual), Straight (clear and direct), Excerpt (editorial style that teases the central idea), and Pithy (lively and compact). Length ranges from 20 to 80 words.
Reviews
There are no reviews for this plugin.
Contributors & Developers
“Kwik AI” is open source software. The following people have contributed to this plugin.
Contributors
Translate “Kwik AI” into your language.
Interested in development?
Browse the code, check out the SVN repository, or subscribe to the development log by RSS.
Changelog
3.1
- New: AI Featured Image generation via FAL.AI, with an editable two-step prompt workflow and preview before applying.
- New: Dedicated Text Model setting for text-only generation.
- New: FAL.AI image-model list populated from the live catalog (cached 12 hours).
- Fix: Custom (OpenAI-compatible) providers can now generate text via
/chat/completionswith a fallback to/api/generate. - Fix: Support newer OpenAI models that reject
max_tokensby retrying withmax_completion_tokens.
3.0
- AI description Gutenberg block with adjustable length and URL-based generation.
- AI description meta box and AI tags meta box improvements.
- Multi-provider support for Ollama, OpenRouter, and OpenAI.