Gemini has officially launched Agentic Trading, making it the first regulated U.S. crypto exchange to allow AI agents to directly execute trades. Unlike traditional API integrations that require manual confirmation for each order, this system grants AI models full access to exchange functions—from querying market data to placing orders—all autonomously.
MCP Architecture: Turning the Full Trading API Into an AI-Readable Interface
Gemini built the system on the Model Context Protocol (MCP), an open protocol led by Anthropic designed to let AI models connect to external tools and services through a unified interface. By integrating its entire trading API into MCP, Gemini enables any MCP-compatible AI model—including Anthropic Claude and OpenAI ChatGPT—to directly access full exchange capabilities. The initial release includes three modular Trading Skills: Get Market Data (real-time price info), Find the Spread (bid-ask spread analysis), and Retrieve Candles (historical candlestick data). These cover everything from simple buy/sell orders to complex multi-leg positions.
Gemini's Bet: AI Handles Execution, Humans Focus on Strategy
In its announcement, Gemini made clear this is more than just a feature launch—it's a paradigm shift: “We believe this is the beginning of a fundamental change in how people interact with financial markets.” The firm added, “Agentic trading isn't just a feature; it's a new paradigm—AI handles execution, patterns, and discipline, while you focus on strategy and goals.” The timing is notable: Gemini cut 25% of its workforce in February, exited the EU, UK, and Australia, and doubled down on the U.S. market. Now with Agentic Trading, it's signaling a pivot from geographic expansion to AI agent infrastructure for the next cycle. Market reaction was muted—GEMI rose only 0.25% to $4.40 on the day, up 7% over the past month but still down 55% year-to-date.
AI × Crypto Bridge Protocols Multiply
Competition in this space is already brewing. Coinbase-incubated x402 protocol, now under the Linux Foundation, targets AI agent on-chain payments and interactions. Stripe-backed Tempo network has launched a Machine Payments Protocol focused on automated settlement between AI agents. By adopting the open MCP standard instead of a proprietary protocol, Gemini ensures Agentic Trading is natively compatible with any MCP-supporting AI model, theoretically enabling faster ecosystem expansion. Whether AI agents become the mainstream interface for retail and institutional order placement, 2026 may be the watershed year.