INDIVIDUAL MCP TOOL

compare_models

Compare two or more models on one task, only within sources that tested every one of them.

aifrontierinstitute.orgnone authenticationAvailability not checked

LIVE ENDPOINT

https://aifrontierinstitute.org/api/mcp

No auth detected

Connect to this endpoint to inspect the live schema for compare_models and invoke it with your own arguments.

{}

Inferred read-only · medium confidence · heuristic, not a guarantee.

No write-capable action terms were found; this is not proof that invocation has no side effects.

CONNECT WITH APPROVAL

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

Codex

~/.codex/config.toml

[mcp_servers.ai-frontier-institute-assurance]
url = "https://aifrontierinstitute.org/api/mcp"
enabled = true

Claude Code

.mcp.json

{
  "mcpServers": {
    "ai-frontier-institute-assurance": {
      "type": "http",
      "url": "https://aifrontierinstitute.org/api/mcp"
    }
  }
}

Claude Desktop

Settings → Connectors → Add custom connector

Name: ai-frontier-institute-assurance
Remote MCP URL: https://aifrontierinstitute.org/api/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": {
    "ai-frontier-institute-assurance": {
      "url": "https://aifrontierinstitute.org/api/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

{
  "servers": {
    "ai-frontier-institute-assurance": {
      "type": "http",
      "url": "https://aifrontierinstitute.org/api/mcp"
    }
  }
}

Generic MCP

Client-specific MCP configuration

{
  "name": "ai-frontier-institute-assurance",
  "transport": "streamable-http",
  "url": "https://aifrontierinstitute.org/api/mcp"
}

MCP Inspector

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

list_tasks — List the business tasks in the AI Frontier Institute's record of independent AI evaluation results, with how many sources and models each covers, when it was last checked, and whether any source measured reliability.
get_task — Get every independent source's results for one task.
get_model — Get every published result for one model across all tasks, grouped by source.
get_input_type_results — Get one task's results for one input type, such as PDF or JPEG on documents: an evidence level (0 to 4) for that type, saying how much independent evidence stands behind it, not how good the models are; then, per test, the most accurate setups, the findings that name the type, and, where the tester published cost from the same run, the most accurate setup, the cheapest within 1 point of the best, and the best at each price.
list_sources — List the sources behind the record: who ran each test, its license and whether we may show its figures, any independence note, and what its score means.