Purpose
ApexVol MCP Server is a Model Context Protocol (MCP) server that gives AI assistants real-time access to ApexVol's options analysis platform. The provider positions it as an entry point to options data and analysis tools for Claude Code, Claude Desktop, claude.ai (remote connector, no installation required), and any MCP-compatible client. The provider states it offers a total of 44 tools and one connector.
Main Capabilities
The provider lists the available tools by category:
- Options Chain (6 tools): get_options_chain, get_expirations, get_options_by_delta, get_stock_price, calculate_expected_move, get_historical_chain.
- Volatility Analysis (7 tools): get_iv_rank, get_volatility_cone, get_volatility_risk_premium, get_term_structure, find_iv_opportunities, get_vix_snapshot, get_monies_surface.
- Greeks and GEX (5 tools): get_gex, get_charm_exposure, get_third_order_greeks, get_greeks_heatmap, get_cross_index_gex.
- Options Flow (3 tools): get_options_flow, get_smart_money_flow, scan_volatility_arb. The provider notes that flow data is end-of-day when the market is closed; if the volume data source is silent, an explanation is returned instead of zero values.
- Strategy Building (5 tools): build_strategy, analyze_strategy, optimize_strategy, simulate_option_chain, calculate_probability_of_profit.
- Risk Management (4 tools): calculate_portfolio_greeks, run_scenario_analysis, generate_stress_tests, get_hedge_recommendations. Positions are passed in as a JSON array (ticker, position_type, quantity, strike, expiration, Greeks); option Greeks must first be obtained via get_options_chain; pure stock portfolios can use plain text (e.g. "AAPL 100 shares").
- Events and Screening (5 tools): get_earnings_calendar, analyze_earnings_history, screen_market, get_market_overview, get_economic_calendar.
- Underlying Analysis (8 tools): get_ticker_analytics (eight views: skew, dividends, borrow_rate, correlation, hv_regimes, price_context, relative_value, greeks_exposure), get_earnings_move_analysis (seven views: mispricing, historical_moves, expected_vs_actual, verdict, seasonality, post_drift, iv_crush), get_max_pain, get_volume_profile, get_zero_dte, get_orats_cores, search_tickers, scan_relative_value.
- Help and Discovery (1 tool): help. help() lists data families and ways to ask questions; help("get_iv_rank"), help("gex"), or help("/iv-rank/{ticker}") returns the corresponding tool, family, or endpoint record; other words perform a search. The provider notes that this tool uses built-in data, does not consume request quota, and is available before setting a token.
The server also registers 41 prompt starters (three per data family, plus one per recipe), which appear in Claude Desktop, Cursor and VS Code's prompt selector.
The provider mentions that version 0.1.4 supports batching: calculate_expected_move and get_iv_rank can take up to 25 tickers at once, get_gex can batch up to 25 tickers, get_options_flow's limit and detail can control row size, and get_zero_dte's strikes_around and detail can control the size of the chain.
How to Use
The provider offers two ways to connect:
Method A — Remote Connector (no installation required) ApexVol is already listed in Claude's connector directory: open https://claude.ai/directory/apexvol,或在 In claude.ai / Claude Desktop, select Customize → Connectors → Discover and search for ApexVol, then select Connect to Claude. Follow the prompts to sign in with your ApexVol account (OAuth); no Python and no configuration file are required. For manual addition, or for other MCP clients, you can use the connector URL:https://apexvol.com/mcp-server(提供商特别注明无尾随斜杠)。Claude Code can also use a remote server, then authenticate when prompted (the /mcp command shows login status).
Method B — Local installation (stdio) The provider recommends installing with pipx so that the apexvol-mcp command is isolated and always on PATH; you can also use a regular pip install apexvol-mcp, or install from source. After installation, you can run apexvol-mcp --check to verify the installation and token authentication; this command prints the client version, authentication result, and remaining rate limits/monthly budget.
Claude Code (CLI) or Claude Desktop must be connected through a configuration file. For Claude Desktop, on macOS edit ~/Library/Application Support/Claude/claude_desktop_config.json, and on Windows edit %APPDATA%\Claude\claude_desktop_config.json. After configuring, restart Claude Code or Claude Desktop to load the MCP server.
Environment variables and authentication Local mode uses token authentication (the APEXVOL_API_TOKEN environment variable; tokens start with avmcp_); the remote connector uses OAuth. The provider states that rate limits are 60 requests per account per minute, 1,000 per hour, and 10,000 per month, shared between local and remote modes.
Use cases
The example queries listed by the provider cover volatility analysis (such as "What's the IV rank for AAPL?", "Show me the volatility cone for SPY"), options chain and Greeks queries, strategy construction (such as iron condor, call spread), risk management (portfolio Greeks, stress testing, hedging suggestions), and market overview and screening. The provider also mentions that the repository includes an Agent Skill for teaching AI assistants how to combine these tools into complete analyses (volatility overvaluation/undervaluation assessment, earnings positioning, seller strategy screening, GEX regime interpretation, strategy stress testing), and includes interpretation thresholds and error handling; when paired withClaude Code, this skill can be copied to the skills directory.
Limitations
- Access requires any paid ApexVol plan (Basic starts at $55 per month). The provider states that each tool returns the same data tier as the web plan, and Pro unlocks the full toolset.
- API and MCP access is included in every paid ApexVol plan; tokens are self-created and managed at apexvol.com/account → API Access.
- The provider notes that the token is sent with every request and should be treated like a password; if leaked, it should be revoked and recreated on the account page.
- Options flow and smart money flow tools return end-of-day data when the market is closed; if the volume data source is silent, an explanation is returned instead of zero values.
- The client is open source (MIT license, see LICENSE), but the provider states that access to the ApexVol platform itself remains subject to subscription restrictions.
Support and documentation
The provider offers the following resources:
- Setup guide:https://apexvol.com/learn/claude-options-data-mcp
- REST API documentation:https://apexvol.com/docs/api
- Single-file API reference (adapted for LLM context windows):https://apexvol.com/docs/api/apexvol-api.md
- Contact email: support@apexvol.com
The conditions, configuration examples, environment variables, and troubleshooting content contained in the repository are organized here only based on the original source text; details not listed are not yet provided.