REST API
Send natural-language questions and structured context to the partner API from any backend or serverless application.
EdgeFinder for Developers
EdgeFinder exposes the same sports research engine that powers its consumer experiences through a partner REST API, JavaScript SDK, CLI, and remote MCP server. Send a question with structured application context and receive analysis grounded in live and historical sports data.
The API, SDK, and MCP server are interfaces to one analysis platform, not separate sports models with different answers.
Send natural-language questions and structured context to the partner API from any backend or serverless application.
Use the typed EdgeFinder client in partner applications for authentication, request construction, conversation state, and error handling.
Connect supported agent clients to EdgeFinder over Streamable HTTP so they can select sports research tools during a conversation.
Research from the terminal, query schedules and standings, inspect odds, or expose EdgeFinder to local MCP clients.
Workflows
Pass roster, betslip, market, portfolio, article, or selected-page context so the response addresses what the user is already viewing.
Call EdgeFinder from an existing application workflow and keep partner identity, end-user identity, and usage isolated.
Give an AI client access to current sports data through MCP instead of asking the model to answer from memory.
Example
The partner API accepts a question plus structured state owned by your product.
POST https://api.edgefinder.io/v1/chat
Authorization: Bearer EF_PLATFORM_API_KEY
Content-Type: application/json
{
"message": "Compare these two trades for my roster",
"league": "nfl",
"endUserId": "partner-user-123",
"context": {
"trades": [{ "receive": ["..."], "send": ["..."] }],
"userState": { "roster": { "players": ["..."] } },
"metadata": { "scoring": "half-ppr", "teams": 12 }
}
}Frequently asked questions
Partner applications integrate against https://api.edgefinder.io/v1/chat. The platform dashboard manages organization keys, usage, and integration settings.
Use MCP when an AI client should discover and call EdgeFinder tools during a conversation. Use the REST API or SDK when your own application controls the request and user experience.
ChatGPT can connect to remote MCP servers directly, without Codex. EdgeFinder is not listed as a supported ChatGPT integration yet because its current MCP authentication uses an EdgeFinder API key, while a public authenticated ChatGPT plugin requires OAuth 2.1 and OpenAI review.
Yes. Partner requests can include structured roster, market, position, portfolio, betslip, page, and user-state context. Send only the fields needed for the requested analysis.
Use EdgeFinder directly, or integrate the sports intelligence platform through the API, SDK, and supported agent interfaces.