---
url: "https://predictiontoolkit.com/tools/pmxt-archive/"
publisher: "PredictionToolkit"
dateModified: "2026-11-03"
---

[← Historical data](<https://predictiontoolkit.com/polymarket-historical-data/>) 

![pmxt archive logo](<https://predictiontoolkit.com/assets/logo-tool-pmxt-archive-61fbcfe4f2b1.webp>)

Dataset

# pmxt archive: recover earlier captures with their version boundaries intact

pmxt archive covers the earlier PMXT V1 and V2 Polymarket capture datasets. Pendulum Flow provides the public mirror, format notes and coverage audit; this archive is separate from PMXT’s current hosted API and SDK.

[**3.5**/ 5Editorial rating](<https://predictiontoolkit.com/tools/pmxt-archive/#reviews>) [Editorial assessment](<https://predictiontoolkit.com/tools/pmxt-archive/#reviews>) 

[Browse the PMXT V2 mirror](<https://archive.pendulumflow.com/browse/pmxt/v2/>) archive.pendulumflow.com

**Pricing:** [See plans and costs ↓](<https://predictiontoolkit.com/tools/pmxt-archive/#pricing>) 

[Screenshots](<https://predictiontoolkit.com/tools/pmxt-archive/#screenshots>) [Explore alternatives](<https://predictiontoolkit.com/tools/pmxt-archive/#alternatives>) 

[Historical data](<https://predictiontoolkit.com/polymarket-historical-data/>) 

## Screenshots and interface

Product preview1 / 2

Mirror index for locating the earlier capture eras.

## What is pmxt archive?

The original archive destination now redirects to general PMXT documentation. Use the explicit V1 and V2 mirror links to locate hourly capture files and read the relevant format before parsing. Parquet, Python and DuckDB belong to the documented data workflow, but a parser still needs to respect the selected version’s schema.

These eras came from earlier single-recorder captures. Missing hours and missed markets require coverage checks, while colliding millisecond timestamps can leave ordering within a millisecond unrecoverable. A checksum can verify a downloaded file against its listed value; the existence of an hourly file does not prove that every market was captured.

### Who it is for

-   Reproducing research built from PMXT archive files.
-   Finding earlier observations of Polymarket books.
-   Comparing data schemas and coverage across capture versions.

## Dataset scope and fields

-   **Capture-era index**
    
    Locate available PMXT V1 and V2 intervals in Pendulum Flow’s mirror index.
    
-   **Public downloads**
    
    Retrieve archived capture files without a trading account.
    
-   **Format by version**
    
    Read the selected era’s format notes before interpreting its rows.
    
-   **File integrity**
    
    Check downloaded files against the checksum list for their mirrored prefix.
    
-   **Coverage audit**
    
    Inspect absent hours and markets omitted by the recorder.
    
-   **Later capture eras**
    
    Use the index to find adjacent later captures while preserving their different formats and collection boundaries.
    

## Data preparation and analysis

### Reproduce an earlier result

Identify whether the analysis used V1 or V2, locate the same hours and verify checksums. Preserve the original schema assumptions so a version change does not silently alter the result.

### Extend a time interval

Read each capture era separately, normalize the fields you need and record boundaries and gaps. Establish compatibility before joining rows from different eras.

## Pricing and total cost

The mirror provides historical capture files without a download charge. It does not represent a hosted API plan: evaluate the current PMXT service's pricing and included access separately from obtaining those older files.

## Dataset access and delivery

1.  Open Pendulum Flow’s archive index.
2.  Select PMXT V1 or V2 and the hours you need.
3.  Read that era’s format notes and coverage audit.
4.  Verify files against the mirrored prefix’s checksum list.
5.  Run a small query for a specific market before processing a long interval.

[Read the PMXT V2 data format](<https://archive.pendulumflow.com/formats/v2>) 

[Browse the PMXT V1 mirror](<https://archive.pendulumflow.com/browse/pmxt/v1/>) 

## Useful links

### What to check

-   Millisecond timestamp collisions can prevent recovery of the event order within a millisecond.
-   An hourly file does not establish complete market coverage.
-   V1, V2 and later eras require version-specific schema handling and explicit boundaries.
-   The original PMXT archive destination redirects to general documentation; use the explicit mirror links for files.
-   The mirrored eras use earlier single-recorder captures, so coverage must be checked for the period and markets you need.

A closer look

## Ratings & reviews

Editorial rating

**3.5**/ 5

0 user reviews

PredictionToolkit editorial

### Our take on pmxt archive

Free versioned captures are useful for exploring earlier market data and reproducing research. Single-recorder gaps, timestamp collisions and changed access links make cleaning and coverage review substantial parts of the work.

**Assessment criteria:** Task usefulness, documented scope, setup clarity, and access and cost transparency.

**Assessment date:** 2026-10-01

**Documentation reviewed:** 2026-10-01

Documentation-based assessment. [How we rate](<https://predictiontoolkit.com/methodology/>) 

### User reviews 0

The main score switches to user ratings after 5 published reviews.

**Be the first to share your experience**

A specific example helps the next person choose the right tool.

## Related tools

[Browse category](<https://predictiontoolkit.com/polymarket-historical-data/>) 

![Pendulum Flow public archive introduction and capture eras](<https://predictiontoolkit.com/assets/cover-pendulum-flow-orderbook-archive-562e9f1691c52d3e-400.webp>)

![Pendulum Flow Orderbook Archive logo](<https://predictiontoolkit.com/assets/logo-tool-pendulum-flow-orderbook-archive-61fbcfe4f2b1.webp>)[Pendulum Flow Orderbook Archive](<https://predictiontoolkit.com/tools/pendulum-flow-orderbook-archive/>) 

Dataset**4.0**Editorial

Inspect hourly Parquet files and coverage audits before reconstructing historical Polymarket depth.

The public mirror, coverage audit and a newer capture era.

![PMXT API entry page with an SDK example](<https://predictiontoolkit.com/assets/cover-pmxt-hosted-api-2a9a2b739dc54d1b-400.webp>)

![PMXT hosted API logo](<https://predictiontoolkit.com/assets/icon-logo-tool-pmxt-hosted-api-636853a0dc32-192-56c21c79cd0be1a3.webp>)[PMXT hosted API](<https://predictiontoolkit.com/tools/pmxt-hosted-api/>) 

Data API**3.5**Editorial

Search and match markets across venues, retrieve history and use supported trading methods with one hosted key.

-   Access API

The current hosted market-data and trading offering.

![PMXT repository overview and supported venue families](<https://predictiontoolkit.com/assets/cover-pmxt-open-source-sdk-8aa7f178a5c5e15f-400.webp>)

![PMXT open-source SDK logo](<https://predictiontoolkit.com/assets/icon-logo-tool-pmxt-hosted-api-636853a0dc32-192-56c21c79cd0be1a3.webp>)[PMXT open-source SDK](<https://predictiontoolkit.com/tools/pmxt-open-source-sdk/>) 

SDK**3.5**Editorial

Use one Python or TypeScript interface across supported prediction venues, with an explicit hosted or local backend choice.

-   Language Python · TypeScript

A client and self-hosted integration toolkit, separate from archive files.

## Frequently asked questions about pmxt archive

### Is this the PMXT SDK?

No. These are historical capture files. The SDK is a separate integration product, and the current hosted API is another offering.

### Where can I find the files?

Pendulum Flow’s archive index links the mirrored PMXT V1 and V2 prefixes.

### Can I join these files directly with V3?

First normalize the schemas and preserve capture-era boundaries and gaps; do not assume the rows are interchangeable.

### Does a free download include current trading services?

No. Free archive downloads do not grant access to PMXT’s current hosted trading API.

In this profile
