This page compares PMXT and ZenHodl for teams evaluating prediction-market data access. PMXT is primarily a unified, open-source trading and data SDK for multiple live venues. ZenHodl is primarily a historical archive vendor with Polymarket and Kalshi order-book depth and trade prints, with Kalshi-side rows additionally carrying capture-time game state. If your goal is live trading or data access across many venues, PMXT may be the better first choice. If your goal is owned historical data for backtesting, ZenHodl may fit better.
At a Glance
| Feature | ZenHodl | PMXT |
|---|---|---|
| Primary focus | Historical archives, Polymarket + Kalshi [1] | Unified real-time trading/data API plus open-source SDK [2] |
| Venue coverage | Polymarket + Kalshi [1] | 15 venues listed, including Polymarket, Kalshi, and Limitless [2] |
| Pricing | One-time archives $9–$199; recurring API/dashboard plans Free/$49/$149/$499 per month; free samples [1][3] | MIT-licensed SDK free/self-hosted; hosted API pricing not published [2] |
| Order-book depth | Historical Kalshi depth + trade prints; top-150-by-volume poll subset, not a full census [1][4] | Unified live trading interface; real-time market data [2] |
| Score-sync | Kalshi-side rows only carry capture-time score/period/clock; Polymarket rows in the same archive do not [5] | Not listed in the public SDK material [2] |
| Format | Parquet/CSV [1] | SDK/API access; self-hosted instances [2] |
| Ownership | Download and own the file [1] | Self-host open-source code, or use hosted API [2] |
| Free tier | Free samples for most datasets [3] | MIT open-source SDK, free [2] |
| Kalshi support | Historical depth + trades [1] | Listed among the 15 supported venues [2] |
When to Pick PMXT
- You need a unified API for live execution or live data across 15 listed venues.
- You prefer self-hosted open-source code under the MIT license.
- Your workflow is real-time order placement or live data ingestion across Polymarket, Kalshi, Limitless, and other supported venues.
- You want a venue abstraction layer rather than venue-specific archive files.
- You are willing to contact PMXT for hosted API pricing or run the SDK yourself.
- You do not need long historical Kalshi order-book depth or capture-time game state.
When to Pick ZenHodl
- You need historical Polymarket or Kalshi data for backtesting.
- You specifically need Kalshi order-book depth and trade prints.
- You want score, period, or clock state captured alongside Kalshi order-book rows (this does not apply to the Polymarket rows in the same archive).
- You prefer a one-time purchase and file ownership over a recurring API dependency.
- You want Parquet/CSV files you can download and use offline.
The Key Differences
1. Real-time venue coverage vs historical depth
PMXT's public material lists 15 venues and a unified trading/data interface — broader live venue coverage than ZenHodl's two-venue archive focus [2]. PMXT centers on live access and SDK use. If you need live breadth across venues, PMXT is the stronger fit; if you need historical depth for Polymarket and Kalshi, ZenHodl is the stronger fit.
2. Open-source self-hosting vs one-time archival purchase
PMXT is MIT-licensed and self-hostable — a real advantage for teams that want to inspect the code, control infrastructure, and avoid per-file purchases [2]. ZenHodl is not open-source; it sells one-time archive files from $9–$199 and also runs recurring API/dashboard plans at Free/$49/$149/$499 per month [1][3]. If you prefer open-source code you can audit and run yourself, PMXT is better. If you prefer a bounded one-time data purchase with published pricing, ZenHodl is better.
3. Score-sync scope and archive limits
ZenHodl's Kalshi-side archive rows include captured game state at capture time; this is capture-time, not event-time, and does not apply to the Polymarket rows in the same archive [5]. The Kalshi depth product itself is a top-150-by-volume poll subset (measured cadence ~31s through 2026-07-11, then ~42–47s from 2026-07-12), not a full market census [4]. PMXT's public SDK material does not list historical score-synced rows at all [2]. ZenHodl archives are downloadable Parquet/CSV files; PMXT runtime is either self-hosted code or a hosted API.
PMXT's open-source contribution is valuable for the live unified-API use case. ZenHodl's contribution is narrower: historical Polymarket and Kalshi archives, with Kalshi-side capture-time game state as one feature among several, subject to the coverage and lag limits above.
Free Sample & Next Steps
Try ZenHodl with a free sample: zenhodl.net/samples. See the full catalog and pricing: zenhodl.net/products.
For Research & Citation
One ZenHodl sample carries a citable DOI today: the free single-game MLB Polymarket–Kalshi matched-book sample is mirrored to Zenodo (10.5281/zenodo.20816908, CC BY-NC 4.0) and linked from the samples page [6]. This DOI covers that specific free sample, not the paid Polymarket & Kalshi order-book archive compared on this page — no DOI or Zenodo reference was found for that archive as of this check.
Sources (checked 2026-09-25)
- zenhodl.net/products — ZenHodl archive catalog and pricing
- github.com/pmxt-dev/pmxt — PMXT README: venues, license, access model (re-fetched 2026-09-25; www.pmxt.dev did not load over HTTPS on this check, so all PMXT facts here come from the GitHub README)
- zenhodl.net/pricing — ZenHodl API subscription tiers
- zenhodl.net/static/samples/kalshi_micro_tape_SCHEMA.txt — Kalshi depth schema and coverage disclosure
- zenhodl.net/static/samples/poly_orderbook_archive_SCHEMA.txt — Polymarket/Kalshi table schema, score-sync scope
- doi.org/10.5281/zenodo.20816908 — Zenodo record for the free MLB matched-book sample