PMXT powers cross-market matching
- On May 20, 2026, PMXT emerged in trading discussions as a unified SDK and router for cross-market matching across prediction venues. - PMXT says its router scans 10-plus venues and links identical markets across Polymarket, Kalshi, Limitless and others through one interface. - PMXT’s docs and GitHub repository show live SDKs, router docs and migration guides for developers evaluating cross-venue deployment.
PMXT is being pitched by developers and traders as a way to collapse fragmented prediction-market infrastructure into one software layer. In a recent social-media thread, traders discussing cross-market arbitrage pointed to PMXT as a unified SDK and matching layer for routing across venues such as Polymarket and Kalshi. PMXT’s own website describes the product as a single API for “every major prediction market,” with cross-venue search, matching, price comparison and trading. Its GitHub repository describes the project more bluntly: “CCXT for prediction markets.” ### What exactly is PMXT selling to trading teams? PMXT’s public documentation says the product combines a unified API, venue-specific trading clients and a router that searches across markets on multiple exchanges. The company’s website says the same SDK is used for trading bots, analytics tools and research systems, and that its matching engine links “identical markets” across Polymarket, Kalshi, Limitless and other venues. The GitHub repository shows PMXT offers Python and TypeScript tooling and positions itself as a drop-in replacement for some older market-access stacks. A migration guide in the repository says developers using Dome API can switch to PMXT through a codemod and a unified interface. ### Why would a cross-market arbitrage desk care about one SDK? Prediction markets are fragmented by venue, contract format and API design. PMXT’s sales pitch is that developers should not have to write and maintain separate code paths for each exchange. Its docs say the router “fans out” one query across venues and returns ranked, unified results, which can then be used for price comparison or order placement. That matters because arbitrage systems compete on both discovery and execution. A desk trying to compare the same event contract across two or more venues needs a fast way to identify equivalent markets, normalize prices and route orders without rewriting the surrounding stack each time a venue changes an endpoint or authentication flow. ### How does the “matching” piece work across different venues? PMXT’s website says its matching engine links the same event across different venues and lets users compare prices where “the same market trades.” The router documentation describes this as a cross-venue intelligence layer for search, matching, hedging and arbitrage detection. That does not mean PMXT is itself an exchange. The public material presents it as infrastructure that sits above exchanges, standardizing market discovery and execution logic. In practice, that means the hard problem is not only sending an order fast, but deciding that a Kalshi contract and a Polymarket contract are close enough in wording, timing and settlement terms to be treated as the same tradeable event. ### Why are traders talking about “control planes” and failover here? The social thread that surfaced PMXT framed the tool as part of a more resilient arbitrage stack, not just a convenience library. That framing matches PMXT’s own product language around one codebase, unified search and a router that can scan more than 10 venues for best pricing. A unified control layer can reduce operational sprawl. Instead of monitoring separate venue adapters, data schemas and routing rules, firms can centralize visibility over search, market mapping and order logic. That can also help with failover if one venue API degrades or if a strategy needs to reroute quickly to another market with comparable exposure. ### What is the trade-off in using an integrated layer like this? PMXT’s documentation emphasizes standardization, but integrated routing layers also concentrate risk. A bug in market matching, symbol normalization or venue selection can propagate across the stack more quickly than in a fully bespoke system. The appeal is still clear in the public materials. PMXT says developers can trade across venues from one codebase, compare prices in real time and build on open-source tooling that is already live on GitHub and PyPI. For teams evaluating it now, the next step is public and immediate: PMXT’s docs, router pages, quick-start guides and repository are already available for testing, with recent commits and package updates posted in May 2026.