SoldFetch

SoldFetch is an independent eBay listing data service for developers, resellers, analysts, and teams building pricing or inventory research tools. It provides sold and active search, category search, and individual item lookup through a shared API contract.

Applications can query supported eBay sites in the United States, United Kingdom, and Germany. Search supports up to 200 results per request. Available fields can include displayed prices, conditions, shipping, and seller details, depending on the listing and source.

Developers can use REST directly or connect an MCP-compatible client. An installable eBay skill provides operation instructions and examples for agent workflows. REST and MCP use the same SoldFetch account and request allowance.

SoldFetch preserves explicit data limitations. Displayed listing prices are not verified settlement amounts. Accepted-offer prices and shipping may be unknown; search sale dates are supplier-reported and unverified. Keep these distinctions when using observations for comparable-price research.

A successful search page or item lookup uses one request. Failed requests are not charged, and successful idempotent replays do not consume another request. Plans have hard request limits without automatic overages.

Who it is for: developers integrating eBay observations into applications; resellers comparing displayed prices and listing conditions; pricing and inventory research teams building repeatable workflows; agent builders needing bounded eBay tools through MCP.

Pricing: Basic is free with 100 units, 10 requests/minute and 1 concurrent request. Starter is $9/month with 2,000 units; Growth is $29/month with 10,000 units; Scale is $79/month with 50,000 units. REST and MCP are included. No card is required for the advertised free signup allowance.

Getting started is straightforward: create a SoldFetch account, receive the free allowance of 100 units, and make a first sold-listing search over REST or connect an MCP-compatible client using the same key. Because plans use hard request limits with no automatic overages, teams can predict cost while prototyping pricing tools, resale research workflows, and agent integrations. Response fields keep price basis and date quality explicit, so downstream tools can avoid treating displayed prices as verified settlement amounts.

Learn more: homepage https://soldfetch.com/ , pricing https://soldfetch.com/pricing , APIs https://soldfetch.com/apis , documentation https://docs.soldfetch.com/ , quickstart https://docs.soldfetch.com/quickstart , MCP guide https://docs.soldfetch.com/guides/ai-agents , MCP endpoint https://soldfetch.com/api/mcp , agent skill https://soldfetch.com/skills/ebay-data/SKILL.md , OpenAPI https://soldfetch.com/api/v1/openapi.json , public examples https://github.com/ricciflow-api/soldfetch-api-examples , MCP metadata https://github.com/ricciflow-api/soldfetch-mcp .

What the product screenshots show: eBay sold listings in one simple API call; ground resale prices in observed listings; from category discovery to listing context; connect eBay data with your AI agents; three marketplaces with one predictable model. All listing values and response excerpts in the gallery are illustrative.

SoldFetch is independent and is not affiliated with eBay. eBay is a trademark of eBay Inc.

Typical use cases include building a repricing assistant that checks recent sold-listing observations, enriching an inventory spreadsheet with item details, comparing the same product across the United States, United Kingdom, and German marketplaces, and giving an AI agent a small set of bounded tools for eBay research instead of unrestricted browsing.

For developers, the documentation includes a quickstart, an OpenAPI description, and public example repositories. For agent builders, the MCP endpoint exposes three tools: search_ebay_sold, search_ebay_category, and get_ebay_item. Authentication uses a SoldFetch key, and the same plan allowance applies whether a request arrives over REST or MCP.

Data quality is treated as part of the product. Each response identifies the basis of a displayed price and marks fields such as accepted-offer amounts, shipping, and sale dates as unknown or unverified when the source cannot confirm them. This helps resale researchers and pricing teams make conservative comparisons instead of overstating certainty.