LLMIntel Model Lifecycle

ai.llmintelv0.2.0Updated Sep 29, 2026

Check whether an LLM model id is deprecated, retiring, or retired, and what to migrate to.

VerifiedStreamable HTTPWeb executableAI & ML

Installation

In SourceWeft

  1. Open LLMIntel Model Lifecycle in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "model-lifecycle": {
      "type": "http",
      "url": "https://llmintel.ai/v1/mcp"
    }
  }
}

README

@llmintel/mcp

An MCP server that tells your coding agent whether a model id is safe to use.

LLMs are trained on a snapshot of the world and will confidently write gpt-4-32k into your code long after it stops answering. This server gives the agent a live lookup for whether a model is deprecated and when it stops working. It returns the replacement too. Answers are normalized across OpenAI, Anthropic, Azure AI Foundry, AWS Bedrock, Google, and Cohere, and parsed from each provider's own deprecation pages.

No API key, no signup. The catalog is public.

[Add to Cursor] [Install in VS Code]

Install

Add it to any MCP host. The package runs straight from npm via npx.

Cursor

In .cursor/mcp.json:

json
{  "mcpServers": {    "llmintel": {      "command": "npx",      "args": ["-y", "@llmintel/mcp"]    }  }}

Claude Code

bash
claude mcp add llmintel -- npx -y @llmintel/mcp

Claude Desktop

Same shape as the Cursor block above, in claude_desktop_config.json.

Hosted endpoint (no install)

The same five tools are served over Streamable HTTP at https://llmintel.ai/v1/mcp. Hosts that take a URL need no Node and no package:

json
{  "mcpServers": {    "llmintel": {      "url": "https://llmintel.ai/v1/mcp"    }  }}

The endpoint is stateless and read-only. It answers from the same catalog the npm package queries.

Tools

ToolReturns
check_modelWhether one model id is safe to use: lifecycle state, the retirement deadline in days, the replacement, and the source link.
list_retiring_modelsWhat breaks in the next 90 days. Past-due models are listed first, then upcoming ones soonest-first.
suggest_replacementThe provider's own recommendation for what to move to. Falls back to same-provider active models when none was published.
search_modelsCatalog search filtered by provider and lifecycle state.
recent_lifecycle_changesThe change feed across all providers, for questions like "what was deprecated this month".

Example

You: Before we ship this, check the model ids in src/agents/.

The agent calls check_model for each one and gets back:

DO NOT USE — this model is retired; API calls to it fail.
"claude-sonnet-4-20250514" resolves to the tracked model anthropic/claude-sonnet-4-20250514.Model: claude-sonnet-4-20250514 (anthropic/claude-sonnet-4-20250514)Provider: anthropicLifecycle state: retired — retired; calls failDeprecated: 2026-04-14 (105 days ago)Retirement: 2026-06-15 (43 days ago)
The provider has not named a replacement. Use suggest_replacement for options.Pricing/limits: $3/1M in · $15/1M out
Source: https://docs.anthropic.com/en/docs/about-claude/model-deprecationsProvider's own term: "Retired"

Deadlines are always given in days, because a model cannot reliably judge whether 2026-07-30 is soon.

Design notes

A failed lookup is never a safety verdict. If the catalog is unreachable, the tool returns an MCP error and says so. An agent that read a network failure as "no deprecation found" would happily ship a retired model id. A model that simply isn't tracked gets the same treatment: it returns "not in the catalog, verify with the provider", never "OK".

Pass whatever string is literally in the code (gpt-4o, anthropic/claude-opus-4-1, azure/gpt-4o) and it resolves to the canonical tracked model.

When the provider's own deprecation notice names a successor, that is what you get. Otherwise the fallback list of same-provider active models is labelled as candidates to evaluate, so an agent can tell the two apart.

Anything past its retirement date is broken now, so it gets its own heading instead of sitting in "retiring soon".

Configuration

Both variables are optional.

VariableDefaultPurpose
LLMINTEL_API_KEYnoneRaises the rate-limit budget. The catalog itself is public, so you do not need this.
LLMINTEL_BASE_URLhttps://llmintel.aiPoint at a self-hosted or staging catalog.

Anonymous callers get 30 requests/minute per IP, enough for interactive agent use.

Data provenance

Every record links to the provider page it was parsed from and preserves the provider's verbatim lifecycle term (sourceTerm), so a normalization decision is always auditable. Changes go through a human verification queue before publication. Collector freshness is public at /v1/status.

The same data is available as a plain REST API, also without a key. See llmintel.ai/docs.

Development

bash
pnpm --filter @llmintel/mcp buildpnpm exec vitest run packages/mcp        # protocol-level tests against a fake catalog
# Drive the built binary against a live catalogpnpm --filter @llmintel/mcp smokeLLMINTEL_BASE_URL=http://localhost:3000 pnpm --filter @llmintel/mcp smoke

License

MIT © LLMIntel

Source: README.md at commit ad9e351

Tools

0
Tool metadata has not been indexed yet.

Version history

1
  1. v0.2.0LatestSep 16, 2026