Gemini has launched Agentic Trading, a new system that allows AI models such as ChatGPT and Claude to connect directly to user trading accounts and execute crypto transactions. The feature runs through the Model Context Protocol (MCP), an open standard that gives AI agents direct API access to external systems.
The exchange describes the rollout as “the first agentic trading tool to be available directly through a regulated U.S.-based exchange.” The system enables AI to place trades, monitor markets, and manage risk based on user-defined strategies.
Gemini integrated its full trading API with MCP, which removes the need for separate custom infrastructure. Users can connect existing AI tools and allow them to act on trading instructions inside the exchange environment.
Trading model shifts from apps to autonomous execution
Agentic Trading changes the interaction model between users and markets. Instead of manually tracking charts or placing orders, users define strategy rules while AI systems handle execution.
Gemini explains that agentic trading means “your AI agent acts on your behalf. That includes placing trades, monitoring markets, and managing risk automatically.”
The system reframes trading as strategy design rather than order execution. Users specify conditions, and AI agents respond through integrated exchange functions.
The exchange also positions this shift as a broader change in market structure. In its announcement, Gemini wrote,
“We believe we're at the beginning of a fundamental shift in how people interact with financial markets.”
MCP integration enables cross-model connectivity
The Model Context Protocol forms the technical base of the system. MCP allows AI models to access external APIs and perform actions inside connected platforms.
Gemini integrated its entire trading infrastructure into MCP. This gives AI agents access to all exchange functions without additional wrappers or custom connectors.
The system supports multiple AI models, including ChatGPT and Claude. Users can select their preferred model and link it directly to trading functions through MCP-compatible connections.
This structure positions AI agents as execution layers between user intent and market action.
Trading Skills define modular automation tools
Gemini introduced Trading Skills as modular functions that AI agents can call during execution. These pre-built tools form the operational layer of the system.
The initial set includes:
- Get Market Data, which retrieves real-time prices and order book information
- Find the Spread, which calculates bid-ask differences across trading pairs
- Retrieve Candles, which provides historical OHLCV data for analysis
These tools allow AI systems to perform structured market analysis and execute trades based on predefined triggers.
The exchange describes Skills as the toolkit that agents use to act. A user can define a condition such as buying Bitcoin at a specific price or reacting to spread changes, and the agent uses Skills to execute those instructions.
Early ecosystem links to broader AI payments infrastructure
The launch arrives alongside wider development of machine-driven financial infrastructure. Protocols such as x402 and the Machine Payments Protocol (MPP) also connect AI systems to financial operations, although they focus on payments rather than exchange execution.
Gemini’s approach centers on trading execution within a regulated exchange environment. This differentiates it from broader machine-to-machine payment systems that operate outside exchange structures.
The system relies on MCP as a shared communication standard across these emerging AI-finance tools.
Financial context surrounds product rollout
Gemini introduced Agentic Trading during a period of financial pressure for the company. The exchange completed an IPO late last year at $28 per share, but the stock has since declined significantly, trading near $4.45 at the time of reporting and dropping below $4 in recent sessions.
The company also went through internal restructuring earlier this year, including workforce reductions and leadership departures across key roles. Gemini reduced operations in several international regions while increasing focus on U.S.-based activity and AI-driven systems.
Despite this, the exchange continues to expand product development. Earlier moves by the founders included large Bitcoin transfers into Gemini hot wallets, reflecting continued activity within core trading infrastructure.
Early technical positioning and market reaction
Gemini frames Agentic Trading as a system that separates strategy from execution. Users define intent while AI handles operational decisions within predefined constraints.
The integration of MCP places Gemini among early financial platforms adopting AI agent infrastructure at the execution level rather than the advisory level.
Market reaction followed quickly after the announcement, with trading activity increasing around Gemini-linked assets and broader attention directed toward AI infrastructure in crypto markets.
Structural shift toward AI-driven market interaction
Gemini’s launch signals a shift in how trading systems interact with automation tools. Instead of using AI for analysis alone, the exchange allows AI systems to operate directly inside trading environments.
The system links strategy definition, market data access, and order execution into a single workflow controlled by AI agents.
Gemini states that this model prioritizes execution efficiency while keeping users in control of strategy design. The long-term direction depends on the adoption of MCP-based infrastructure across trading platforms and the continued integration of AI models into financial systems.
Matvii Diadkov
Disclaimer: All materials on this site are for informational purposes only. None of the material should be interpreted as investment advice. Please note that, despite the nature of much of the material created and hosted on this website, HODL FM operates as a media and informational platform, not a provider of financial advisory services. The opinions of authors and other contributors are their own and should not be taken as financial advice. If you require advice, HODL FM strongly recommends contacting a qualified industry professional.
