ROIC.ai has released a complete technical guide detailing its financial data API and its new Model Context Protocol (MCP) server. The platform covers over 68,000 tickers across 70+ global exchanges, providing everything from historical stock prices to earnings call transcripts. This documentation is essential for developers building financial analysis tools or integrating AI agents that require real-time market data.

The Core API and Data Coverage

The REST API at api.roic.ai is structured around four main data groups: Fundamentals, Earnings, Market Data, and Company Data. It offers over 40 years of financial statement history and 17+ years of earnings call transcripts. Authentication is handled via Bearer tokens in production, while query parameters are supported for quick testing. The guide emphasizes that supplying both authentication methods simultaneously results in a 401 Unauthorized error.

Pricing Tiers and Limitations

ROIC.ai offers four pricing plans: Free, Individual ($29/month), Professional ($89/month), and Enterprise (Custom). The Free plan allows 5 requests per minute and 2 years of historical data. Individual users get 300 requests per minute and 5 years of data, while Professional and Enterprise plans offer unlimited requests and full access to 40+ years of history. Crucially, commercial redistribution rights are exclusive to the Enterprise tier.

The 36-Tool MCP Server

The standout feature is the hosted MCP server at mcp.roic.ai, which exposes 36 read-only tools to AI assistants. This allows Large Language Models (LLMs) to fetch data using natural language queries without custom code. For example, asking "What did Apple's management say about AI in their latest earnings call?" triggers the get_latest_earnings_call tool. The server supports clients like ChatGPT, Claude, and VS Code, with specific setup instructions provided for each.

Key Takeaways

  • Massive Coverage: Access to 68,000+ tickers and 70+ exchanges with up to 40 years of historical data.
  • AI-Ready Integration: The 36-tool MCP server enables direct LLM access to financial metrics without custom API wrappers.
  • Clear Pricing Tiers: Free to Enterprise plans, with commercial rights restricted to the top tier.
  • Standardized Responses: Consistent JSON error envelopes and pagination structures make integration predictable.

The Bottom Line

ROIC.ai’s combination of deep historical data and a standardized MCP interface removes the traditional friction of building financial AI agents. While the Enterprise-only commercial license is a hurdle for startups, the free tier’s MCP access provides an immediate, low-code path for developers to prototype data-driven assistants.