TradingView enables AI agent connectivity via MCP server in public beta

TradingView just opened a door that a lot of traders have been waiting for. The charting platform launched a public beta of its Model Context Protocol (MCP) server on September 16, letting paid subscribers connect AI agents directly to their accounts for live market data, watchlist management, and natural-language analysis.

The integration works with AI assistants like Anthropic’s Claude, turning what used to require manual chart-reading and screen-hopping into something closer to a conversation. Ask your AI assistant about a stock’s fundamentals, and it pulls the data straight from TradingView. No API keys required.

How the MCP server actually works

MCP, or Model Context Protocol, is a standardized way for AI models to interact with external tools and data sources. The connection happens through a single server URL (https://mcp.tradingview.com/mcp), authenticated via OAuth 2.1, which means users log in once and the AI handles the rest.

The toolset available through the beta is already fairly broad. Users can pull real-time quotes, access historical price data, run screeners, manage watchlists and alerts, and review company fundamentals, news, and regulatory filings. All through plain English queries rather than clicking through menus.

There are guardrails in place. The beta enforces a rate limit of roughly 100 tool calls per minute, and daily request caps may apply depending on usage. Market data delivered through the MCP server also comes with delays, which is worth noting for anyone hoping to use this for split-second trading decisions.

Access is limited to users on Essential-tier subscriptions and above. Trial accounts don’t qualify, which effectively gates the feature behind TradingView’s paid tiers.

Part of a broader AI push

This isn’t TradingView’s first foray into AI-powered features. Back in April 2026, the platform rolled out AI Chart Copilot, a tool designed to assist users with technical analysis directly within the charting interface. The MCP server takes that concept further by making the entire TradingView data layer accessible to external AI agents rather than confining intelligence to a single built-in feature.

Community developers had previously built unofficial MCP servers for TradingView. The official beta makes those workarounds largely unnecessary. It also introduces something the community versions couldn’t offer: direct account-linked interactions, meaning AI assistants can manage actual watchlists and alerts, not just read publicly available data.

TradingView is framing itself as the “data and tools layer” for AI-driven market analysis. The company is actively soliciting feedback through an official Reddit thread.

What this means for the trading platform landscape

Retail brokers and trading platforms across the industry have been racing to integrate AI capabilities throughout 2026. TradingView’s MCP approach is architecturally different: instead of building a chatbot inside the platform, it’s turning the platform into a resource that any sufficiently capable AI can tap into.

The delayed market data limitation is the most obvious area to watch. Professional traders and active day traders need real-time feeds, and the MCP server’s delayed quotes constrain its utility for time-sensitive decision-making.