LLMIntel MCP server

This is a small helper that lets your AI look up whether a model name (like gpt-4o or claude-sonnet-4) is still alive, about to be switched off, or already dead. It is handy if you or your team write code that calls AI models and you want to avoid shipping something that will break next month. You do not need an account or a key to use it.

What this MCP server does

You ask your AI something like whether a model name in your code is still fine. Your AI sends that name to this helper. The helper checks the LLMIntel catalog, which is built from each provider's own deprecation pages. It comes back with the model's status, the retirement date in days, the suggested replacement, and a link to the provider's page. Your AI then explains the answer to you in the chat.

Click to zoom

What you can do with it

• Check whether a single model name is safe, retiring, or already retired
• List which models stop working in the next 90 days
• Ask what to migrate to when a model is going away
• Search the catalog by provider or by status
• See what changed recently across all providers
• Spot retired model names hiding in your code before you ship

Try asking your AI

• “Is gpt-4-32k still safe to use, or has it been retired?”
• “Check every model name in src/agents/ and tell me which ones are dead.”
• “What models are retiring in the next 90 days?”
• “What should I move to instead of claude-sonnet-4-20250514?”

What it gives back to you

You get back a short answer in the chat: the model's status (active, deprecated, retiring, or retired), how many days until it stops working, the replacement if the provider named one, and a link to the source page. For list questions, you get a sorted list, with already-broken models first. If the catalog cannot be reached, it says so instead of pretending everything is fine.

Before you start

Who it's for

Developers and anyone who writes or reviews code that calls AI models, plus teams trying to keep their model names up to date.