mcpdirectory.dev
9,129 servers · updated 23 Sept

Model Lifecycle MCP server

community Active last commit 56 d
{
  "mcpServers": {
    "llmintel": {
      "command": "npx",
      "args": [
        "-y",
        "@llmintel/mcp@latest"
      ],
      "env": {
        "LLMINTEL_API_KEY": "<llmintel_api_key>",
        "LLMINTEL_BASE_URL": "https://..."
      }
    }
  }
}
file .mcp.json (in your project root)install npx -y @llmintel/mcpversion 0.2.0

What it does

This server checks whether LLM model IDs are deprecated, retiring, or retired across OpenAI, Anthropic, Azure, AWS Bedrock, Google, and Cohere. It returns the lifecycle state, retirement deadline, and recommended replacement model. It is for developers who need to verify that model IDs in their code are still supported before deployment.

  • Check individual model IDs to confirm they are not deprecated or retired before shipping code.
  • Find what breaks in the next 90 days to plan migration timelines.
  • Search the catalog by provider and lifecycle state to identify active alternatives.

Tools this server exposes

ToolDoes
check_modelReturns the lifecycle state, retirement deadline in days, replacement model, and source link for a single model ID.
list_retiring_modelsReturns models that will stop working in the next 90 days, sorted by retirement date.
suggest_replacementReturns the provider's recommended replacement model, or active alternatives from the same provider if no replacement was published.
search_modelsSearches the catalog filtered by provider and lifecycle state.
recent_lifecycle_changesReturns the change feed across all providers to identify recently deprecated or retired models.

Before you enable it

An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.