Binance launches Agent OS to connect AI agents to financial infrastructure

Binance launches Agent OS to connect AI agents to financial infrastructure

Users can connect with agents through applications such as ChatGPT, Claude Code, Codex and Cursor.

Binance has launched Agent OS as a standardized developer layer for connecting AI applications and agents to its financial infrastructure, the company said Thursday.

According to Binance, the platform gives developers access to trading, market data, wallets, payments and onchain capabilities spanning both crypto and traditional markets, and forms part of Binance Intelligence, the exchange’s strategic effort around AI-powered financial products.

Built for AI developers, fintech companies and quantitative traders, Agent OS aims to address the fragmentation developers face when building agentic finance applications. The system supports both Binance-built integrations and custom AI agents.

Advertisement

“It gives everyone from developers to quantitative traders the reliable data, low‑latency infrastructure, and standardized interfaces they need to deploy AI‑driven strategies,” Jeff Li, VP of Product at Binance, stated.

Compatible AI tools including ChatGPT, Claude Code, Codex and Cursor can be authorized to retrieve market data, view account information and perform supported trading operations. Users determine the permissions and limits available to each agent and can place individual agents in dedicated subaccounts to segregate their funds and trading activity.

The infrastructure combines existing Binance APIs with the Wallet Agentic Hub, x402 programmable payment capabilities and Skill Hub, while adding support for MCP. By using MCP, an open standard for connecting AI applications to external tools, Binance aims to give developers a common interface for building agents that can interact with its various products without having to create separate integrations for each one.

According to the team, the first MCP implementation provides access to market data, read-only account information and trading functionality. Users can assign an agent to a dedicated subaccount, set its permissions and revoke access whenever they choose.

Agents can see balance, portfolio information and transaction history within the designated subaccount, along with balance and portfolio information from the main account. They are blocked from accessing personal account information such as email addresses and KYC data.

The architecture also separates trading oversight from agent intelligence, Binance noted. The exchange retains visibility into activity conducted through its infrastructure, including the resulting orders, and applies its existing controls to that activity.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
Binance launches Agent OS to connect AI agents to financial infrastructure
Binance launches Agent OS to connect AI agents to financial infrastructure

Users can connect with agents through applications such as ChatGPT, Claude Code, Codex and Cursor.

Share

Add us on Google

Binance has launched Agent OS as a standardized developer layer for connecting AI applications and agents to its financial infrastructure, the company said Thursday.

According to Binance, the platform gives developers access to trading, market data, wallets, payments and onchain capabilities spanning both crypto and traditional markets, and forms part of Binance Intelligence, the exchange’s strategic effort around AI-powered financial products.

Built for AI developers, fintech companies and quantitative traders, Agent OS aims to address the fragmentation developers face when building agentic finance applications. The system supports both Binance-built integrations and custom AI agents.

Advertisement

“It gives everyone from developers to quantitative traders the reliable data, low‑latency infrastructure, and standardized interfaces they need to deploy AI‑driven strategies,” Jeff Li, VP of Product at Binance, stated.

Compatible AI tools including ChatGPT, Claude Code, Codex and Cursor can be authorized to retrieve market data, view account information and perform supported trading operations. Users determine the permissions and limits available to each agent and can place individual agents in dedicated subaccounts to segregate their funds and trading activity.

The infrastructure combines existing Binance APIs with the Wallet Agentic Hub, x402 programmable payment capabilities and Skill Hub, while adding support for MCP. By using MCP, an open standard for connecting AI applications to external tools, Binance aims to give developers a common interface for building agents that can interact with its various products without having to create separate integrations for each one.

According to the team, the first MCP implementation provides access to market data, read-only account information and trading functionality. Users can assign an agent to a dedicated subaccount, set its permissions and revoke access whenever they choose.

Agents can see balance, portfolio information and transaction history within the designated subaccount, along with balance and portfolio information from the main account. They are blocked from accessing personal account information such as email addresses and KYC data.

The architecture also separates trading oversight from agent intelligence, Binance noted. The exchange retains visibility into activity conducted through its infrastructure, including the resulting orders, and applies its existing controls to that activity.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.