One query fans out to Facebook Marketplace, eBay, Depop, and Poshmark simultaneously — and your AI merges, dedupes, and ranks what comes back.
The secondhand market is fragmented on purpose: the same vintage jacket might be $45 on Depop, $70 on Poshmark, and $30 (plus a drive) on Facebook Marketplace. Whoever checks the most marketplaces wins — but nobody wants to run the same search four times with four different filter UIs.
Cross-marketplace search fixes the mechanics. With Secondhand MCP connected to Claude or ChatGPT, a single "search all" query runs against Facebook Marketplace, eBay, Depop, and Poshmark simultaneously. The AI gets structured results from each — titles, prices, conditions, locations, photos — and does what AIs are good at: merging, comparing, and telling you which listing is actually the best deal.
Individual marketplace searches are available from the free tier up. The one-shot "search all" fan-out is a Pro+ ($10/mo) feature — on lower tiers you can still ask the AI to search marketplaces one at a time. Deep-research tools (search/fetch), which also search all accessible marketplaces per query, are metered on every tier.
The AI sees all results with titles, photos, and prices, so it can spot and flag cross-posted items — one of the things a language model is genuinely better at than a keyword engine.
One search-all query costs four marketplace searches' worth of infrastructure (including headless-browser scraping for Depop and Poshmark), so it lives on the top tier. Pro users can still search marketplaces individually in sequence.
Yes — the server exposes the standard search/fetch tools Deep Research expects, and those tools search all accessible marketplaces per query on any tier.
Facebook Marketplace, eBay (official API, BYOK), Depop, and Poshmark today. Local marketplaces are the lane we're expanding along.
Free tier included. No credit card required.