Meet WhiteBIT AI Hub: Trade, Analyze, and Build with AI
AI is changing how people interact with markets. Instead of constantly switching between charts, account pages, documentation, and code, traders and developers can manage more of their workflows through natural-language commands.
WhiteBIT AI Hub brings this experience to WhiteBIT.
It is a set of tools that allows users to interact with WhiteBIT market data, account functions, and trading infrastructure through locally installed AI clients and developer environments. Developers can also build automated workflows and AI agents without creating a custom API integration from scratch.
WhiteBIT AI Hub is available today and includes an open-source MCP server, a command-line interface, AI-ready documentation tools, and prebuilt analytical Skills.
Trade by Talking, Not Clicking
The WhiteBIT MCP server allows users to connect the WhiteBIT API to a supported AI client and communicate with the exchange in plain language.
At launch, authenticated WhiteBIT API functionality can be connected to Claude Code or Codex installed locally on the user’s device.
Once the connection is configured, users can ask the AI to:
- check the current price or order book for a trading pair;
- show spot balances or open collateral positions;
- list active and historical orders;
- place, modify, or cancel an order;
- transfer assets between WhiteBIT account balances;
- analyze market data before making a decision.
The MCP server connects the locally installed AI client to WhiteBIT tools for market data, spot and collateral trading, wallet operations, lending, sub-accounts, and other exchange functions.
This is not a separate trading simulator or a limited AI sandbox. Authenticated commands are executed through the WhiteBIT API, giving users access to real exchange functionality from a supported local AI environment.
Direct trading and account management through the standard ChatGPT chat interface are not available yet.
Orders That Keep Watching the Market
Active traders cannot stay in front of a chart around the clock. WhiteBIT AI Hub helps users translate a trading plan into an order without manually navigating several exchange screens.
Through the MCP server, a supported AI client can work with regular and conditional order types, including limit, market, stop, OCO, stop-limit, and trigger-market orders.
For example, a user can ask the AI to prepare an order that activates only after a specific market condition is reached.
Once placed, the order logic is handled by the exchange. The user does not need to keep the AI client open or continuously monitor the chart for the trigger.
AI does not replace the trader’s judgment. It provides another interface for expressing and executing a decision. Users should review order parameters before submitting them, apply appropriate risk controls, and give API keys only the permissions required for the intended task.
One Hub, Four Ways to Work with AI
WhiteBIT AI Hub combines tools for traders, developers, analysts, and teams building automated systems.
1. MCP Server: Natural-Language Access to WhiteBIT
The WhiteBIT MCP server connects Claude Code or Codex running locally on a user’s device to the WhiteBIT API.
After configuration, users can make requests in everyday language instead of manually constructing API calls. The AI identifies the appropriate WhiteBIT tool, prepares the request, and returns the result inside the local AI environment.
Public market information, including prices, tickers, order books, and market statistics, can be accessed without a WhiteBIT account or API key.
Trading, balances, positions, and other private account functions require:
- a WhiteBIT account;
- a configured API key;
- Claude Code or Codex installed locally;
- the WhiteBIT MCP server connection.
The MCP server is open source, allowing developers to inspect its implementation, contribute to the project, report issues, and build agent-based workflows on top of it.
2. WhiteBIT CLI: Exchange Tools in the Terminal
For developers who prefer commands over conversations, the WhiteBIT CLI brings trading and account management into the terminal.
It supports market data, spot and collateral trading, lending, account operations, sub-accounts, and other parts of the WhiteBIT API. The CLI also provides:
- JSON output for scripts and automation tools;
- multiple credential profiles;
- a dry-run mode for previewing requests;
- shell completion;
- support for automated workflows and CI/CD pipelines.
Public market-data commands work without authentication, while private account and trading commands require API credentials.
This removes much of the repetitive work previously required to create and maintain a custom WhiteBIT API integration. Developers can use ready-made commands in scripts, internal tools, monitoring systems, or automated trading workflows.
3. AI-Ready Documentation Without Setup
Not every user needs to install a trading tool or create an API key.
Every page in the WhiteBIT API documentation includes an AI menu. From it, users can:
- copy the page in an AI-friendly format;
- view its raw Markdown;
- open the documentation in ChatGPT, Claude, Perplexity, or Grok;
- access available MCP connection options.
This feature helps users explore and understand the documentation with an AI assistant without setting up the trading MCP server.
Opening a documentation page in ChatGPT or another AI chat does not connect that chat directly to a WhiteBIT account and does not enable trading. It only provides the selected documentation as context for the conversation.
This is the simplest entry point to WhiteBIT AI Hub: no account, API key, or local installation is required to explore the documentation with AI.
4. Skills: Ready-Made Market Analysis
WhiteBIT AI Hub also includes Skills: predefined workflows that a supported AI client can run on demand.
They are designed for tasks such as:
- market intelligence using macroeconomic, on-chain, and institutional data;
- technical analysis with more than 20 indicators, including EMA, RSI, and MACD;
- sentiment analysis and fear-and-greed signals;
- portfolio and trade reviews;
- algorithmic order planning.
Instead of writing a long prompt or manually combining several data requests, users can ask for the required analysis in plain language and let the relevant Skill guide the workflow.
Open and Local-First
The WhiteBIT trading MCP server runs locally on the user’s device.
API credentials are used to authenticate requests to WhiteBIT and are not stored by WhiteBIT AI Hub. The server’s source code is publicly available on GitHub, allowing developers to review how the integration works.
Users remain in control of the permissions assigned to each API key. A read-only key can be used for account queries, while trading permissions should be enabled only when needed.
Withdrawal permissions are not required for trading and should remain disabled unless a specific use case requires them.
For additional separation, users can create a dedicated sub-account and API key for AI-assisted workflows.
Built for Traders and Developers
WhiteBIT AI Hub supports several levels of interaction.
For traders, it provides a conversational way to research markets, inspect an account, and manage orders through Claude Code or Codex running locally.
For analysts, it brings market data and prebuilt analytical workflows into a supported AI environment.
For developers, it offers an open MCP layer and a full-featured CLI for building scripts, applications, and automated agents.
For teams, it makes WhiteBIT documentation easier to use with AI and provides structured JSON output for internal tools and CI/CD workflows.
Users can start with public market data, explore the documentation through the AI menu, or configure authenticated tools when they are ready to work with their WhiteBIT account.
How to Get Started?
WhiteBIT AI Hub tools are free to use. Standard WhiteBIT fees apply when an order or another fee-based exchange operation is executed.
To explore the available tools and choose the right setup, visit the WhiteBIT AI Hub.
Developers can open the AI documentation hub, follow the MCP server and CLI setup guides, or explore the open-source MCP server on GitHub.
Install Claude Code or Codex locally, connect the WhiteBIT MCP server, configure an API key, and start interacting with WhiteBIT through natural-language commands.
Explore the WhiteBIT AI Hub and discover a new way to work with the exchange through AI.
