Model ID Cheatsheet

About

AI coding agents often struggle with outdated or incorrect model names, leading to hallucinations and broken code. This tool acts as a critical external knowledge base, offering instant, accurate access to API model IDs, pricing, context windows, and capabilities for 46 models across 7 leading AI providers. By integrating with the Model Context Protocol (MCP), it empowers agents to look up verified information before generating code, ensuring they always use the correct, most current, and cost-effective models. It is built in Go, distributed as a lightweight 10MB binary, requires no external calls, offers sub-millisecond responses, and automatically updates weekly.

Key Features

• 0 GitHub stars
• Provides accurate API model IDs, pricing, and specs for 46 models across 7 providers.
• Offers 6 dedicated AI agent tools: get_model_info, list_models, recommend_model, check_model_status, compare_models, and search_models.
• Lightweight Go-based server with a single 10MB binary, sub-millisecond responses, and weekly auto-updates.
• Supports direct MCP connection and various coding agent integrations (Claude Code, Cursor, Windsurf, Codex CLI, OpenCode).
• Includes built-in security features like rate limiting, connection limits, and input sanitization.

Use Cases

• Ensuring AI agents generate correct model API calls to prevent using deprecated or non-existent model IDs.
• Helping agents identify the most suitable and cost-effective models for specific tasks based on capabilities and budget.
• Enabling agents to compare various AI models side-by-side based on pricing, context window, and capabilities.