Question 1

What is Telonex?

Accepted Answer

Telonex is the industry standard for historical prediction market data. It provides clean, tick-level trades, order book updates, quotes, and onchain settlement data for Polymarket — the world's largest prediction market on the Polygon blockchain. With the most complete historical dataset available for Polymarket (20 billion+ data points across 450,000+ markets), Telonex serves traders, researchers, and data scientists who need reliable, analysis-ready prediction market data without the overhead of running their own data infrastructure.

Question 2

What prediction market data does Telonex provide?

Accepted Answer

Telonex provides six data channels for Polymarket: (1) Trades — every executed trade with timestamp, price, size, and side; (2) Quotes — best bid/ask snapshots at every change; (3) Order Book Snapshots — full depth at 5, 25, or all price levels; (4) Onchain Fills — blockchain settlement data with maker/taker wallet addresses and USDC amounts; (5) Markets — metadata for all 450,000+ markets including titles, outcomes, and resolution status; (6) Tags — category labels for filtering markets by topic.

Question 3

What is tick-level prediction market data?

Accepted Answer

Tick-level data means every individual event is recorded — every trade execution, every order book update, every quote change. This is the highest resolution of market data available, as opposed to aggregated candle/OHLCV data that summarizes activity into time intervals. Tick-level data is essential for accurate backtesting, market microstructure research, and understanding how information flows through prediction markets.

Question 4

How far back does Polymarket historical data go?

Accepted Answer

Onchain fills data covers the full history (3+ years) of Polymarket settlements on the Polygon blockchain from market inception. Off-chain data (trades, quotes, order book snapshots) is collected daily and coverage is growing — check the free markets dataset for exact availability dates per market and channel.

Question 5

How many markets does Telonex cover?

Accepted Answer

Telonex covers all 450,000+ Polymarket markets, totaling over 20 billion data points. This includes active markets, resolved markets, and historical markets across every category — politics, sports, crypto, finance, science, culture, and more. The markets dataset, which is free to download without an account, lists every market with its metadata.

Question 6

What is onchain fills data?

Accepted Answer

Onchain fills are the actual settlement records from the Polygon blockchain for every Polymarket trade. Each fill includes the maker wallet address, taker wallet address, token IDs, price, size, transaction hash, and USDC amount. This data enables wallet-level analysis like tracking whale activity, calculating per-trader P&L, and monitoring smart money flows. In the Parquet files, each trade is mirrored into two rows (one in each outcome token's partition with complementary prices), so filter to mirrored=false to avoid double-counting volume.

Question 7

What blockchain does Polymarket run on?

Accepted Answer

Polymarket operates on the Polygon blockchain (previously Matic Network), an Ethereum Layer 2 scaling solution. All Polymarket trades are settled onchain as conditional token transfers using the Gnosis Conditional Token Framework. Telonex collects and normalizes this blockchain data into clean, analysis-ready Parquet files through the onchain fills channel.

Question 8

How is Telonex different from Polymarket's own API?

Accepted Answer

Polymarket's CLOB API is designed for trading — placing orders, checking positions, and streaming live data. It is not designed for bulk historical data retrieval and has rate limits that make downloading large datasets impractical. Telonex is purpose-built for historical data analysis: we provide pre-cleaned, daily-partitioned Parquet files that you can download in bulk via our API or Python SDK. No pagination, no rate limit juggling, no WebSocket management — just download the data and analyze.

Question 9

What format is the data delivered in?

Accepted Answer

All data is delivered in Apache Parquet format, the industry standard for analytical workloads. Parquet files are columnar, compressed, and optimized for fast reads in tools like pandas, Polars, DuckDB, Spark, and R. A single Parquet file is typically 5-50x smaller than the equivalent CSV, and loads 10-100x faster for analytical queries. Our Python SDK can convert directly to pandas DataFrames with a single function call.

Question 10

How do I download Polymarket historical data?

Accepted Answer

There are two ways to download data from Telonex. (1) Python SDK: Install the telonex package with "pip install telonex[dataframe]" and call get_dataframe() with your API key to download data directly into a pandas DataFrame. (2) REST API: Make HTTP GET requests to the download endpoint with your API key, exchange, channel, market, and date range parameters. Both methods return Parquet files partitioned by date for efficient downloads. See our documentation at telonex.io/docs for full details.

Question 11

Can I use Telonex data with pandas and Python?

Accepted Answer

Yes. Telonex is designed for Python-first workflows. Install our SDK with "pip install telonex[dataframe]" and use get_dataframe() to download data directly as a pandas DataFrame. The SDK handles authentication, caching, and file management automatically. You can also load Parquet files directly with pandas.read_parquet(), or use Polars and DuckDB for faster processing of large datasets.

Question 12

Is there a Python SDK?

Accepted Answer

Yes. The Telonex Python SDK is available on PyPI. Install it with "pip install telonex" for core download functionality, or "pip install telonex[dataframe]" for pandas DataFrame support. The SDK provides simple functions to download data, check availability, and list datasets. It handles file caching locally so repeated queries for the same data are instant. Full SDK documentation is at telonex.io/docs/sdk/python.

Question 13

Can I access the data without writing code?

Accepted Answer

The free markets and tags datasets can be accessed by anyone via direct Parquet URLs without writing code or creating an account. For the full historical data (trades, quotes, order books, onchain fills), you need an API key and either our Python SDK or REST API. We do not currently offer a web-based data download interface, but you can browse sample data on our datasets page at telonex.io/datasets.

Question 14

Is any data available for free without an account?

Accepted Answer

Yes. The markets and tags datasets are completely free and require no login or API key. The markets dataset contains metadata for all 450,000+ Polymarket markets including titles, outcomes, resolution status, and data availability dates. You can download it directly with pandas: pd.read_parquet("https://api.telonex.io/v1/datasets/polymarket/markets"). Additionally, the free trial includes 5 file downloads across all data channels.

Question 15

Can I backtest prediction market strategies with Telonex data?

Accepted Answer

Yes. Telonex provides the tick-level granularity required for meaningful strategy backtesting on prediction markets. You can download every trade, every order book snapshot, and every quote change for any market, then simulate your strategy against actual historical execution data. This includes realistic slippage modeling using order book depth data, something impossible with OHLCV candles alone. See our use cases page for Python code examples.

Question 16

Can I use Telonex data for academic research?

Accepted Answer

Yes. Many researchers use Telonex data to study prediction market efficiency, information aggregation, forecasting accuracy, and market microstructure. Our data is research-grade: deduplicated, normalized, and delivered in Parquet format that works with any statistical computing environment. The free markets dataset lets you explore what's available before subscribing. For academic inquiries about bulk access or special pricing, contact us at support@telonex.io.

Question 17

How do I track whale activity on Polymarket?

Accepted Answer

Use the onchain fills channel, which includes maker and taker wallet addresses for every trade. You can group fills by wallet address to calculate each trader's total volume, position size, and P&L. To identify whales, sort traders by cumulative volume or position size. To track specific wallets over time, filter the fills data by address and join with market metadata to see what they're trading and when.

Question 18

Can I calculate P&L for Polymarket traders?

Accepted Answer

Yes, using onchain fills data. For each wallet, you need to track three components: USDC flows (money in/out of trades), YES token position changes, and NO token position changes. Final P&L is calculated as: total USDC flows + (YES position * resolution value for YES) + (NO position * resolution value for NO). Note that cross-token trades (where maker_asset_id differs from taker_asset_id) represent about 44% of election market trades and must be handled correctly.

Question 19

Can I reconstruct the order book from historical data?

Accepted Answer

Yes. Telonex provides order book snapshots at three depth levels: top 5 levels, top 25 levels, and full depth. Each snapshot captures the complete state of the order book at that moment, so you can reconstruct the book at any point in time without replaying incremental updates. For most backtesting use cases, the 25-level snapshots provide sufficient depth for slippage modeling.

Question 20

How often is the data updated?

Accepted Answer

Off-chain data (trades, quotes, order book snapshots) is updated daily, typically available by early morning UTC for the previous day's activity. Onchain fills data covers the full history from market inception. The free markets and tags datasets are also updated daily. All data is partitioned by date, so you can incrementally download only new data as it becomes available.

Question 21

How does Telonex handle data quality?

Accepted Answer

Telonex applies multiple data quality measures: event deduplication to remove duplicate records, integrity checks to verify data completeness, normalized schemas across all channels, and redundant data collection using multiple independent collection instances. Data is delivered in daily partitions with consistent column types and naming conventions. We also provide detailed schema documentation for every field in every channel.

Question 22

Does Telonex deduplicate trades?

Accepted Answer

Yes. Telonex deduplicates events at the collection layer using multiple redundant instances to ensure no trades are missed and no duplicates are included. Each trade appears exactly once in the trades channel. For onchain fills, each blockchain trade emits multiple events: an OrderFilled for the maker, an OrderFilled for the taker (with the exchange contract address as one of maker/taker), and an OrdersMatched event. Telonex uses the maker's OrderFilled as the primary record — ignoring the taker's OrderFilled to avoid double-counting — and joins OrdersMatched to add taker fields (taker address, taker_asset_id, taker_side), producing one combined row per trade. The Parquet files then mirror each row into the sibling asset partition (with 1 minus the original price), so filter to mirrored=false to get one row per trade.

Question 23

What is mirroring in onchain fill data?

Accepted Answer

The blockchain emits multiple events per trade, but Telonex combines them into a single row. The Parquet files then mirror each row into two: one in the maker's asset partition (mirrored=false) and one in the sibling (complementary) asset partition (mirrored=true) with the complementary price (1 minus the original price). This ensures each outcome token's partition is self-contained. Telonex includes the "mirrored" field in every onchain fills record so you have the full information.

Question 24

How much does Telonex cost?

Accepted Answer

Telonex offers three tiers: Free Trial ($0) with 5 file downloads to evaluate the data, Plus ($69/month) with unlimited downloads and full historical data for personal use, and Enterprise (custom pricing) with dedicated support, custom integrations, and commercial use license. The free markets and tags datasets are always free for everyone, no account needed.

Question 25

Is there a free trial?

Accepted Answer

Yes. Sign up with Google (no credit card required) and you get 5 free file downloads across all data channels — trades, quotes, order book, and onchain fills. This lets you evaluate the data quality and schema before subscribing. The markets and tags datasets are always free without any account.

Question 26

How does billing work?

Accepted Answer

Subscription is billed monthly through Stripe. You can cancel anytime from your dashboard and retain access until the end of your billing period. There are no annual commitments or long-term contracts.

Question 27

What is the difference between Plus and Enterprise?

Accepted Answer

Plus ($69/month) is for individual traders and researchers who need unlimited data access for personal use. Enterprise is for teams and companies that need custom data delivery, multiple API keys, dedicated support, Point-in-Time accuracy guarantees, and a commercial use license. Contact sales@telonex.io for Enterprise pricing.

Question 28

Do you offer refunds?

Accepted Answer

We offer a free trial so you can evaluate the data before subscribing. If you have concerns about your subscription, contact us at support@telonex.io and we will work with you to find a solution.

Question 29

What are the API rate limits?

Accepted Answer

The Telonex API is designed for bulk data downloads, not high-frequency real-time queries. Rate limits are generous for typical data analysis workflows. If you need higher throughput or custom delivery, contact us about Enterprise access. Full API documentation is available at telonex.io/docs/api/overview.

Question 30

How large are the data files?

Accepted Answer

File sizes vary by market activity and data channel. A single day of trades for an active market is typically 1-10 MB in Parquet format. Order book data is larger due to depth information. The Python SDK downloads and caches files locally, so repeated access to the same data is instant. Parquet compression typically achieves 5-50x reduction compared to equivalent CSV data.

Question 31

What tools work with Parquet files?

Accepted Answer

Parquet is supported by virtually every modern data tool: Python (pandas, Polars, PyArrow), R (arrow package), SQL engines (DuckDB, Spark, BigQuery, Snowflake, Athena), Julia, Rust, Go, and more. For quick exploration, DuckDB can query Parquet files directly with SQL without loading them into memory. Our Python SDK provides the simplest path — one function call to get a pandas DataFrame.

Question 32

Does Telonex provide real-time data?

Accepted Answer

No. Telonex is focused on historical data for backtesting, research, and analysis. Off-chain data is updated daily. If you need real-time Polymarket data, you can use Polymarket's own CLOB API for live trading and streaming, and Telonex for historical analysis and backtesting.

Frequently Asked Questions

About Telonex & the Data

Data Access & Formats

Use Cases

Data Quality & Coverage

Pricing & Billing

Technical

Ready to get started?