Conversation
supportWhen you access models via the model maker, such as OpenAI or Anthropic, AI Engine does not support the latest models after release. It always needs an update of AI Engine. Currently, AI Engine does not support Claude Sonne 4.6 through the Anthropic API. This doesn’t make sense. I use numerous other AI tools, and they all support the AI developers’ models immediately after release via the API, without requiring an update to the AI tool. AI Engine supports the latest models via OpenRouter, but many of these models lack some features available through the developers’ API. For example, web search or MCP cannot be used this way. AI Engine requires a feature to refresh the model list for all model providers. This can be implemented very simply.
Hi @michael8888 , I understand your frustration, but I believe there are some significant technical misunderstandings in your message that I’d like to clarify 😉 You mentioned that it doesn’t make sense to wait for an update, but from a developer’s perspective, it is the only way to ensure the quality you expect. To address your point about the Anthropic Model List (API): while it returns a list of names, it does not return pricing, context window limits, or feature support. As you have previously mentioned in other threads, pricing accuracy is vital. If I implemented the simple refresh you suggest, AI Engine would show new models instantly, but with completely incorrect pricing and metadata . I choose not to release broken or unverified model configurations to my users. Regarding your comment on OpenRouter: it actually does not natively support MCP. They provide workarounds , but not direct API integration for it. My goal is to make sure that when a feature is added to AI Engine, it is done properly and natively so that it is a long-lasting solution for your workflow. Now, about the timeline: – Claude Sonnet 4.6 Release: 3:00 AM UTC – AI Engine Update: ~9 AM UTC A 6-hour turnaround for a stable, tested, and price-accurate integration is actually very fast. I would appreciate it if, instead of assuming these things are simple or senseless , you could look a bit deeper into the technical constraints we manage to keep the tool reliable for you, and discuss them nicely with the developers. To me, it’s very important to keep an agreeable atmosphere where we don’t point fingers negatively 🌴😌 I am sincerely doing my best to AI Engine, and I hope you can see the value in a tool that prioritizes being “right” over being “first by five hours”. This reply was modified 3 months, 1 week ago by Jordy Meow . Reason: Link to https://platform.claude.com/docs/en/api/models/list This reply was modified 3 months, 1 week ago by Jordy Meow . Reason: Actually took me 6 hours, not 5 hours ;)
Jordy, when I mentioned it doesn’t make sense, I didn’t mean to offend. If it truly doesn’t make sense, then it simply doesn’t. There’s no reason to sugarcoat it. I’ve said before that AI Engine is the best WordPress plugin ever, and these minor flaws won’t change that. Pricing is indeed important. However, last time I complained, the issue was that the pricing was off by orders of magnitude. I don’t mind if it’s just a few cents or dollars off. The pricing of successor models is typically similar to that of their predecessors. Given how quickly the AI industry is evolving, hardcoding pricing information and metadata indeed makes little sense. AI Engine is the only AI tool I know that works this way. You can store this metadata in a central database and have AI Engine retrieve it from there. An AI agent can easily keep the database up to date. You don’t want to do such things manually in 2026. I see you’ve just updated the AI Engine. Thanks for that!
Hi @michael8888 , I appreciate the follow-up and the kind words about the plugin. Regarding the idea of a central metadata database, I have considered it, but I decided against it. Moving metadata to a central server would make every WordPress installs dependent on an external system. In the WordPress ecosystem, users generally prefer autonomy. Also, if a central DB managed by an agent were to have an issue or a hallucination, it could instantly affect thousands of sites! 😅 By keeping this data within the plugin, I ensure there is no middleman that can fail. Regarding the API endpoints, OpenRouter actually provides full details because we worked together in the past to ensure their metadata is comprehensive (somehow, I am actually the one who pushed them to add the metadata/details in their API). Unfortunately, I haven’t had that same level of communication with OpenAI or Anthropic. I would much prefer for them to improve their own endpoint to include pricing and features natively, as that is where the fix truly belongs rather than in a third-party database. In the meantime, I believe this little manual check is the best way to keep the tool open, predictable, and reliable for everyone.
These days, it’s quite common for complex WordPress plugins to provide a backend service. As long as you don’t force users to sign up, that’s fine. Make this feature optional and store the metadata locally as a fallback. Jensen Huang recently claimed that the models no longer hallucinate. While this is certainly an exaggeration, I can confirm that, after decades of working with countless software engineers, humans hallucinate much more than AI does. In fact, the idea that AI hallucinates more than humans is a hallucination in itself. For instance, if you point out to a software engineer that their application has flaws or bugs, they will hallucinate 90% of the time, whereas a model will admit its mistakes in 99% of cases. 😉 I wouldn’t expect reliable standards from inference providers anytime soon. The competition is fierce, and the industry is advancing quickly. To keep up with developments, you have to build your own online service. I am confident that AI agents can update the database more reliably than humans.
Hi @michael8888 , I understand your frustration, but I believe there are some significant technical misunderstandings in your message that I’d like to clarify 😉 You mentioned that it doesn’t make sense to wait for an update, but from a developer’s perspective, it is the only way to ensure the quality you expect. To address your point about the Anthropic Model List (API): while it returns a list of names, it does not return pricing, context window limits, or feature support. As you have previously mentioned in other threads, pricing accuracy is vital. If I implemented the simple refresh you suggest, AI Engine would show new models instantly, but with completely incorrect pricing and metadata . I choose not to release broken or unverified model configurations to my users. Regarding your comment on OpenRouter: it actually does not natively support MCP. They provide workarounds , but not direct API integration for it. My goal is to make sure that when a feature is added to AI Engine, it is done properly and natively so that it is a long-lasting solution for your workflow. Now, about the timeline: – Claude Sonnet 4.6 Release: 3:00 AM UTC – AI Engine Update: ~9 AM UTC A 6-hour turnaround for a stable, tested, and price-accurate integration is actually very fast. I would appreciate it if, instead of assuming these things are simple or senseless , you could look a bit deeper into the technical constraints we manage to keep the tool reliable for you, and discuss them nicely with the developers. To me, it’s very important to keep an agreeable atmosphere where we don’t point fingers negatively 🌴😌 I am sincerely doing my best to AI Engine, and I hope you can see the value in a tool that prioritizes being “right” over being “first by five hours”. This reply was modified 3 months, 1 week ago by Jordy Meow . Reason: Link to https://platform.claude.com/docs/en/api/models/list This reply was modified 3 months, 1 week ago by Jordy Meow . Reason: Actually took me 6 hours, not 5 hours ;)
Jordy, when I mentioned it doesn’t make sense, I didn’t mean to offend. If it truly doesn’t make sense, then it simply doesn’t. There’s no reason to sugarcoat it. I’ve said before that AI Engine is the best WordPress plugin ever, and these minor flaws won’t change that. Pricing is indeed important. However, last time I complained, the issue was that the pricing was off by orders of magnitude. I don’t mind if it’s just a few cents or dollars off. The pricing of successor models is typically similar to that of their predecessors. Given how quickly the AI industry is evolving, hardcoding pricing information and metadata indeed makes little sense. AI Engine is the only AI tool I know that works this way. You can store this metadata in a central database and have AI Engine retrieve it from there. An AI agent can easily keep the database up to date. You don’t want to do such things manually in 2026. I see you’ve just updated the AI Engine. Thanks for that!
Hi @michael8888 , I appreciate the follow-up and the kind words about the plugin. Regarding the idea of a central metadata database, I have considered it, but I decided against it. Moving metadata to a central server would make every WordPress installs dependent on an external system. In the WordPress ecosystem, users generally prefer autonomy. Also, if a central DB managed by an agent were to have an issue or a hallucination, it could instantly affect thousands of sites! 😅 By keeping this data within the plugin, I ensure there is no middleman that can fail. Regarding the API endpoints, OpenRouter actually provides full details because we worked together in the past to ensure their metadata is comprehensive (somehow, I am actually the one who pushed them to add the metadata/details in their API). Unfortunately, I haven’t had that same level of communication with OpenAI or Anthropic. I would much prefer for them to improve their own endpoint to include pricing and features natively, as that is where the fix truly belongs rather than in a third-party database. In the meantime, I believe this little manual check is the best way to keep the tool open, predictable, and reliable for everyone.
These days, it’s quite common for complex WordPress plugins to provide a backend service. As long as you don’t force users to sign up, that’s fine. Make this feature optional and store the metadata locally as a fallback. Jensen Huang recently claimed that the models no longer hallucinate. While this is certainly an exaggeration, I can confirm that, after decades of working with countless software engineers, humans hallucinate much more than AI does. In fact, the idea that AI hallucinates more than humans is a hallucination in itself. For instance, if you point out to a software engineer that their application has flaws or bugs, they will hallucinate 90% of the time, whereas a model will admit its mistakes in 99% of cases. 😉 I wouldn’t expect reliable standards from inference providers anytime soon. The competition is fierce, and the industry is advancing quickly. To keep up with developments, you have to build your own online service. I am confident that AI agents can update the database more reliably than humans.