It has been quite some time since the CyberSEO Pro and RSS Retriever plugins for WordPress received updates to their AI image generation features, which are in high demand in the auto-blogging industry. Meanwhile, AI technology is constantly evolving, and image models are advancing just as rapidly as language models.
Today, we are pleased to announce that both of our autoblogging plugins now support the long-awaited GPT-Image-2 model from OpenAI. In professional circles, it is currently praised above all for its phenomenal accuracy in rendering text. Whereas previously the letters in generated images resembled incomprehensible Latin characters from a parallel universe, the model now renders complex signs, labels, and even entire paragraphs with nearly 100% accuracy.
For owners of news blogs that aggregate and automatically rewrite content, this opens up enormous possibilities. The new GPT-Image-2 model allows users to generate not just abstract images, but fully-fledged, unique illustrations where the article’s headline or a key phrase can be seamlessly integrated into the graphic itself against an attractive background. This guarantees that every news item is visually complete and entirely unique for search engines and readers alike.
Unfortunately, all the advantages of GPT-Image-2 are negated by a serious drawback stemming from OpenAI’s approach to “security,” which has already crossed all boundaries of common sense. The company requires users to undergo mandatory identity verification through the Persona service, providing their passport details or ID. Such requirements from a private web resource seem absurd, especially given the legislation of European Union countries, where the unjustified collection of biometric data and documents constitutes a direct violation of privacy rights.
Now, OpenRouter comes to the rescue once again, allowing you to use the capabilities of GPT-Image-2 (also known as GPT-5.4 Image 2) without any verification. In our previous article, we already explained why having a single key for all tasks is convenient, but now it’s also a matter of your anonymity. To spare you from the questionable need to present your passport to verifiers, we’ve ensured full compatibility of CyberSEO Pro and RSS Retriever with all of the hub’s image engines. Our implementation allows you to seamlessly use any current and future service models while maintaining confidentiality and detailed cost control in a single admin panel.
In addition to models from OpenAI, the OpenRouter interface now offers an impressive list of alternatives that is constantly expanding. You can use the Flux.2 versions of Klein 4B, Max, Flex, and Pro from Black Forest Labs; the powerful Gemini 2.5 Flash, Gemini 3.1 Flash Image, and Gemini 3 Pro Image from Google; as well as the promising Riverflow V2 in Fast, Standard, Max, and Pro variants. ByteDance Seedream 4.5 and the convenient Auto Router feature for automatic model selection are also available.
All of these AI-powered image models are now available in automatic mode via the feed settings (the Media enrichment tab) and for more precise content customization in the HTML post template editor. A new shortcode [or_image] was added to handle images, similar to the [or_text] shortcode you’re already familiar with. This unification allows you to embed unique images created by top-tier neural networks anywhere in your article. Specify the necessary parameters in the template to achieve this.
OpenRouter is more than a way to bypass absurd verification requirements. It’s the most effective tool for business. A single account covers your needs for both text and image models. You will get transparent expense tracking and detailed usage statistics right in the service’s admin panel. This is often cheaper and more convenient than working with each provider directly.
WordPress automatic blogging with Midjourney
