AI & Machine Learning

models.rip

An MCP server for exploring models.rip, an archive of retired AI models and dead benchmarks with details on graves, providers, and departed models.

2 endpoints8 known toolsFirst detected September 11, 2026Last detected September 11, 2026

ENDPOINT 1

https://models.rip/mcp

No auth detected

MCP server metadata

Name: models-rip
Version: 1.0.0

Capabilities

tools

Server instructions

models.rip is an archive of AI models that are no longer served. Use it to answer what happened to a specific model and when, what replaced it, which dated snapshot ids a model absorbed, and which models are scheduled to be shut off next. Every date is the one the provider itself published, and every grave cites the provider page it came from. Call cemetery_overview first if you have not used this server before. Use search_graves when you do not know the exact slug — it matches retired snapshot ids too, so a model id someone pinned in production will find its grave. Use get_grave for the full record. An absence is an answer: a model with any snapshot still answering in its maker's own lineup is not buried here, and neither are fine-tunes, rehosts or reseller listings. Do not use this server to check whether an API is up right now, or to compare models that are alive.

Known tools 4

search_graves

Find retired AI models and dead benchmarks buried at models.rip.

Inferred read-only

get_grave

The full record of one grave at models.rip: its epitaph, its launch and retirement dates, the shutdown date its provider published, its successor, its pricing and context window, and the provider pages every date is cited from.

Inferred read-only

list_plots

Every provider with a plot at models.rip and how many models each has buried, largest first, plus the benchmarks plot.

Inferred read-only

cemetery_overview

What this archive is, what it deliberately leaves out, how many graves it holds, and where the raw dataset lives.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.models-rip]
url = "https://models.rip/mcp"
enabled = true

Claude Code

.mcp.json

{
  "mcpServers": {
    "models-rip": {
      "type": "http",
      "url": "https://models.rip/mcp"
    }
  }
}

Claude Desktop

Settings → Connectors → Add custom connector

Name: models-rip
Remote MCP URL: https://models.rip/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "models-rip": {
      "url": "https://models.rip/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

{
  "servers": {
    "models-rip": {
      "type": "http",
      "url": "https://models.rip/mcp"
    }
  }
}

Generic MCP

Client-specific MCP configuration

{
  "name": "models-rip",
  "transport": "streamable-http",
  "url": "https://models.rip/mcp"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

ENDPOINT 2

https://models.rip/.well-known/mcp

No auth detected

MCP server metadata

Name: models-rip
Version: 1.0.0

Capabilities

tools

Server instructions

Known tools 4

search_graves

Find retired AI models and dead benchmarks buried at models.rip.

Inferred read-only

get_grave

Inferred read-only

list_plots

Every provider with a plot at models.rip and how many models each has buried, largest first, plus the benchmarks plot.

Inferred read-only

cemetery_overview

What this archive is, what it deliberately leaves out, how many graves it holds, and where the raw dataset lives.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.models-rip]
url = "https://models.rip/.well-known/mcp"
enabled = true

Claude Code

.mcp.json

{
  "mcpServers": {
    "models-rip": {
      "type": "http",
      "url": "https://models.rip/.well-known/mcp"
    }
  }
}

Claude Desktop

Settings → Connectors → Add custom connector

Name: models-rip
Remote MCP URL: https://models.rip/.well-known/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "models-rip": {
      "url": "https://models.rip/.well-known/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

{
  "servers": {
    "models-rip": {
      "type": "http",
      "url": "https://models.rip/.well-known/mcp"
    }
  }
}

Generic MCP

Client-specific MCP configuration

{
  "name": "models-rip",
  "transport": "streamable-http",
  "url": "https://models.rip/.well-known/mcp"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.