Limited time offer: 6 months freeClaim
EdgeFinder Logo
EdgeFinder
Back to Integrations

EdgeFinder for ChatGPT

Bring live NFL, NBA, MLB, and Polymarket data into ChatGPT. Use the EdgeFinder CLI locally, or call the EdgeFinder API directly from a custom GPT or agent workflow.

Summary

ChatGPT on its own does not know tonight's injury report or the current Polymarket line. EdgeFinder gives it a research layer that does.

The easiest path is the EdgeFinder CLI on your machine, which you can pipe into ChatGPT or use from a local agent. You can also integrate EdgeFinder into a custom GPT via API.

Use cases

  • - Player prop research grounded in real data rather than training cutoffs.
  • - Odds movement checks that reference the current market rather than stale numbers.
  • - Injury impact analysis using the latest reports.
  • - Multi-leg parlay reasoning with per-leg probability context.

Setup

  1. Step 1

    Install the EdgeFinder CLI

    The CLI gives you an authenticated command-line tool that can answer betting questions against live data.

    npm i -g @edgefinder/cli
  2. Step 2

    Authenticate

    Log in with a magic link or set an API key from your integrations page.

    edgefinder login
  3. Step 3

    Query EdgeFinder, paste into ChatGPT

    Run an EdgeFinder query in your terminal and paste the structured answer into ChatGPT for further reasoning or formatting.

    edgefinder ask "Saquon Barkley rushing yards tonight"
  4. Step 4

    Or call the API from a custom GPT

    Expose EdgeFinder as an action in a custom GPT using the API and your API key. Custom GPTs can call EdgeFinder for any research step.

Example queries

  • - Get me Polymarket pricing for tonight's NBA games and flag any gaps with sportsbook spreads.
  • - Pull the last five games for Tyler Herro and summarize usage trends.
  • - Analyze Magic vs 76ers and compare Polymarket and sportsbook prices.
  • - Find prop lines where the market looks slower than the injury news.

FAQ

Does ChatGPT support MCP?

ChatGPT does not currently ship native MCP support. The EdgeFinder CLI is the simplest path today, with custom GPTs and API calls as an alternative.

Can I call EdgeFinder from my own agent code?

Yes. The EdgeFinder API works from any language and any agent framework, not just official MCP clients.