cryo + Chainstack: A developer's guide to blockchain data mastery with Python


In blockchain data exploration, we previously introduced you to cryo, Paradigm's powerful command-line interface tool. As you might recall, this tool is a beacon for developers, researchers, and blockchain enthusiasts, optimizing the process of extracting data from various blockchain networks. Our initial journey through cryo revealed its data formatting efficiency and seamless integration with Chainstack Global Elastic Nodes.


Learn how to use cryo and how it works with cryo: Your gateway to blockchain data.

Now, we embark on a sequel, bridging cryo with the world of Python. This guide will show you how to use the Python wrapper made for the cryo CLI by covering setup, basic usage, and data extraction and manipulation using common Python libraries.

Python and cryo for blockchain data manipulation

Python is known for its simplicity and data manipulation and analysis capability. The Python wrapper allows you to couple cryo's Rust-based efficiency for data extraction with Pythonโ€™s data manipulation capabilities. This integration enhances the analytical power at your fingertips, allowing you to leverage Python's rich library ecosystem for in-depth data analysis, visualization, and machine learning.

Prerequisites and setup

This section will lay the groundwork for integrating the cryo tool with Python. This process involves ensuring that your system has the necessary tools and libraries and installing the Python wrapper for cryo.


Before diving into the installation process, ensure your environment is primed for the task. The following prerequisites are essential:

  • Chainstack GEN RPC: Get a high-performance Chainstack Global Elastic Node RPC before starting.

Follow these steps to deploy an Ethereum node:

  1. Sign up with Chainstack.
  2. Deploy a node.
  3. View node access and credentials.

To follow this guide, deploy a Standard Ethereum node, which will default to a Global Elastic Node.


Please note that GENs are only available as a full node at the moment, archive GENs coming soon.

Once you deploy the node, you'll have access to an RPC endpoint, which will look like this:

Create a .env file in your root directory and place the endpoint in it.

  • Rust: Rust must be installed in your system for cryo to work, the Python integration is a lightweight wrapper for the cryo CLI, so youโ€™ll still need to meet the appโ€™s requirements.


Install Rust following the rustup instructions.

  • Python Environment: Ensure that you have Python installed on your system and create a new virtual environment in your projectโ€™s directory; you can run the following:

    python3 -m venv cryo-and-python

    Then activate the virtual environment with:

    source cryo-and-python/bin/activate
  • Required Libraries: cryo_python depends on several libraries, make sure to install the following libraries,

    pip install maturin pandas polars pyarrow python-dotenv web3 matplotlib

    Note that the python-dotenv web3 matplotlib libraries are not strictly required to run cryo_python, but weโ€™ll use them along the guide.

Installation and Setup

With the prerequisites in place, letโ€™s move on to the installation steps:

  1. Clone the cryo Repository: Use git to clone the cryo repository from GitHub. If you donโ€™t have git installed, you can download it from git.

    git clone
  2. Navigate to the Python Directory:

    cd cryo/crates/python
  3. Build cryo_python:

    • Run the maturin build command:
      maturin build --release
    • This command will compile the Rust code and create a wheel file (.whl) for the Python package.
  4. Install the Python Wrapper:

    • Find the .whl file generated by maturin. It will be located in the target/wheels directory.
    • Install the wheel file using pip:
      pip install --force-reinstall <PATH_TO_WHEEL_FILE>.whl
    • Replace <PATH_TO_WHEEL_FILE> with the actual path to the .whl file generated, it will look like this:

Your current draft provides a solid foundation. To enhance it, we can add more context and details based on the source files, particularly focusing on the functionality and technical nuances of cryo.collect() and cryo.freeze(). Here's an improved version:

Basic Usage of cryo_python

cryo_python serves as a lightweight wrapper for the cryo CLI offers a seamless Python interface to the powerful CLI commands. With cryo_pythonusers can access two principal functions that mirror their CLI counterparts:

  • cryo.collect() extracts blockchain data and returns it as a Python-friendly data frame, enabling direct use within scripts for real-time analysis and manipulation.
  • cryo.freeze() fetches data and saves it to a file, facilitating subsequent use or long-term storage.


Explore the source files for cryo.collect() and cryo.freeze() in the GitHub repository.

cryo.collect() Main Aspects

  1. Asynchronous Support: cryo.collect() includes both async_collect and collect methods, designed to operate asynchronously. This feature is vital for efficiently handling large datasets or high-throughput tasks, ensuring optimal resource utilization and performance.
  2. Multiple Output Formats: cryo.collect() allows you to organize data in various Python-friendly formats for diverse scenarios:
    • Polars DataFrame: Ideal for high-performance data manipulation, leveraging its fast, efficient data handling capabilities.
    • Pandas DataFrame: Provides broad compatibility with Python's extensive data analysis ecosystem.
    • List of Dictionaries: Facilitates easy handling of JSON-like data structures, simplifying serialization.
    • Dictionary of Lists: Offers an alternative structured data format suitable for specific data processing requirements.

cryo.freeze() Main Aspects

  1. Data Type Flexibility: cryo.freeze() can handle single and multiple data types, showcasing its versatility in accommodating various data collection needs.
  2. Argument Parsing: Echoing cryo.collect(), cryo.freeze() also parses additional keyword arguments (*kwargs), enhancing the customization possibilities in data collection and storage.

Usage examples

Having grasped the basics of cryo_python, let's get into practical examples to demonstrate its usage. Throughout this guide, we'll consistently retrieve the RPC endpoint from a .env file.


Ensure you have your RPC endpoint details in a .env file for these examples.

cryo.collect basic example

Start by creating a file named and paste the following code:

import os
import cryo
from dotenv import load_dotenv

# Load environment variables from .env file

# Retrieve the Ethereum RPC URL from environment variables
eth_rpc = os.getenv("ETH_RPC")

# Collect blockchain data using the cryo library and return it as a pandas DataFrame
# Specifying blocks range and output format
data = cryo.collect(

# Displaying the column names of the DataFrame
print("Columns in the DataFrame:")
for column in data.columns:

# Print the entire DataFrame

Here's an explanation of how it works and what it does:

  1. Environment Setup:
    • The code starts by importing the necessary modules: os for environment variable management, cryo for accessing blockchain data, and load_dotenv from the dotenv package to load environment variables from a .env file.
    • It then loads the environment variables using load_dotenv(), which reads the .env file and sets the variables.
  2. Accessing Ethereum RPC Endpoint:
    • The ETH_RPC variable, which contains the URL to an Ethereum RPC endpoint, is fetched from the environment variables using os.getenv("ETH_RPC").
  3. Data Collection with cryo.collect:
    • The cryo.collect function has specific parameters to fetch data from the Ethereum blockchain.
    • datatype: Set to "blocks", indicating that the function should fetch data about blockchain blocks.
    • blocks: Specifies the range of blocks to fetch data for (in this case, from block 18734050 to 18735050).
    • rpc: The Ethereum RPC endpoint URL, passed as eth_rpc.
    • output_format: Set to "pandas", indicating that the data should be returned as a Pandas DataFrame.
    • hex: The boolean parameter set to True will return the data already converted to hexadecimal.
  4. Output:
    • The fetched data is stored in the variable data, a Pandas DataFrame.
    • The script then prints the column names of the DataFrame to provide an overview of the data structure.
    • Finally, it prints the DataFrame data, showing the fetched blockchain data.

The result of this script is a detailed listing of data for the specified range of Ethereum blocks. The DataFrame columns represent each block's attributes, such as block_hash, author, block_number, gas_used, extra_data, timestamp, base_fee_per_gas, and chain_id.

Here is an example of the output in the console:

Columns in the DataFrame:
                                            block_hash                                      author  ...  base_fee_per_gas  chain_id
0    0xdf6d5d7526eb50e68278998b2cc7a519a4c3daddb14a...  0xcdbf58a9a9b54a2c43800c50c7192946de858321  ...       37583327088         1
1    0x2818389bb471ebe60a74fb1865574c0ac50f40daf575...  0xdafea492d9c6733ae3d56b7ed1adb60692c98bc5  ...       38781853745         1
2    0xd65229e3d67c28f71b978ae789df6ee58c27420f8a35...  0x1f9090aae28b8a3dceadf281b0f12828e676c326  ...       38448400269         1
3    0x218f26e524d889d20604ad01b91feec5c2285dc3e747...  0x4675c7e5baafbffbca748158becba61ef3b0a263  ...       38316617096         1
4    0xace296b1c263fee4f35d831086c93aa820577dcc1bea...  0x388c818ca8b9251b393131c08a736a67ccb19297  ...       37638298982         1
..                                                 ...                                         ...  ...               ...       ...
995  0xc59d9ad3444b0352c36f3ec7e3a3561bbc90d9118232...  0x690b9a9e9aa1c9db991c7721a92d351db4fac990  ...       71896254942         1
996  0xd611033a7769913ba0e8abdc8ae0ab0fee224435c512...  0x4838b106fce9647bdf1e7877bf73ce8b0bad5f97  ...       70283122265         1
997  0x2a4448fe72e37c169868e37ea4fce71789c31ecc9108...  0x95222290dd7278aa3ddd389cc1e1d165cc4bafe5  ...       73132855124         1
998  0x6ce066a304e334c6a98154b51f2a1b24edf749467424...  0x4838b106fce9647bdf1e7877bf73ce8b0bad5f97  ...       74400095143         1
999  0xbc293dc0e8d0a61f24256433f7faaf2a8e754a5557d9...  0x4838b106fce9647bdf1e7877bf73ce8b0bad5f97  ...       73090669589         1

[1000 rows x 8 columns]

Running this Python script is the equivalent of running this command from the cryo CLI directly:

cryo blocks --blocks 18734050:18735050 --rpc YOU_CHAINSTACK_GEN


Please note that Chainstack endpoints on the Developer plan are limited to 30 RPS, so you might need to add rate limiting to your code; starting from the Growth plan, there is no rate limit.

To manage rate limits, cryo.collect can be adjusted using the requests_per_second parameter:

data = cryo.collect(

cryo.freeze basic example

The principle of cryo.freeze is quite similar to cryo.collect. In a new file, paste this code:

import os
from dotenv import load_dotenv
import cryo

# Load environment variables from the .env file

# Retrieve the Ethereum RPC URL from environment variables
eth_rpc = os.getenv("ETH_RPC")

# Fetch and save blocks data in JSON
data = cryo.freeze(

This script uses cryo.freeze to fetch and save the same block data as a JSON file in the specified directory. The logic and syntax closely follow the cryo CLI. The result is a JSON file containing data for the blocks in the root/blocks_data/ directory.

Since both cryo.freeze and cryo.collect are just wrappers around the CLI; you can use the same commands. Letโ€™s explore a few more examples.

Fetching ERC-20 balances with cryo

This section will guide you in using cryo_python to retrieve ERC-20 token balances from specified addresses and contracts. Weโ€™ll get the balance of the APECoin token in the Binance address in a range of 10,000 blocks.

Start by creating a new Python file and paste the following code:

import os
from dotenv import load_dotenv
import cryo

# Load environment variables

# Access Ethereum RPC URL from environment variables
eth_rpc = os.getenv("ETH_RPC")

# Fetch ERC-20 token balances for a specified address within a block range
data = cryo.freeze(

Executing this script will generate a JSON file containing the ERC-20 balance data structured as follows:

Schema for ERC-20 Balances
- block_number: uint32
- erc20: hex
- address: hex
- balance_binary: binary
- balance_string: string
- balance_f64: float64
- chain_id: uint64

This structure, erc20_balancesefficiently organizes ERC-20 balances by block, offering a clear and accessible format for data analysis.


Check the cryo documentation to find what other datasets you can fetch.

Fetch and manipulate blockchain data

Having explored the basic functionality of cryo_python, let's now get into a more advanced application by integrating it with essential Python libraries for data manipulation and visualization.

Find the top 10 block authors

In this example, we'll fetch Ethereum blockchain data and visualize the top block authors using cryo_python, pandas, and matplotlib.

In a Python file, paste the following:

import os
import time
import pandas as pd
import matplotlib.pyplot as plt
from dotenv import load_dotenv
from web3 import Web3
import cryo

# Constants

# Load environment variables

# Function to get the block range
def get_block_range(web3, lookback_blocks):
    latest_block = web3.eth.block_number
    start_block = max(0, latest_block - lookback_blocks)  # Avoid negative block numbers
    return f"{start_block}:{latest_block}"

# Function to fetch data from cryo
def fetch_block_data(block_range, eth_rpc):
        # Start timer for fetching blocks
        fetch_start_time = time.time()

        # Fetch the block data
        data = cryo.collect(
            "blocks", blocks=[block_range], rpc=eth_rpc, output_format="pandas", hex=True

        # Calculate and print the time taken to fetch the blocks
        fetch_time = time.time() - fetch_start_time
        print(f"Time taken to fetch blocks: {fetch_time:.2f} seconds")
        return data
    except Exception as e:
        print(f"An error occurred while fetching block data: {e}")
        return pd.DataFrame()  # Return an empty DataFrame on error

# Function to plot the top authors
def plot_top_authors(data, num_entries):
    top_authors = data['author'].value_counts().head(TOP_AUTHORS_COUNT)
    plt.figure(figsize=(12, 6))
    plt.xticks(rotation=45, ha='right')
    plt.title(f'Top {TOP_AUTHORS_COUNT} Authors by number of blocks mined from past {num_entries} blocks', fontsize=14)
    plt.xlabel('Author', fontsize=14)
    plt.ylabel('Number of Blocks', fontsize=14)

# Main execution
def main():
    eth_rpc = os.getenv(ETH_RPC_VAR)
    if not eth_rpc:
        raise ValueError(f"Environment variable {ETH_RPC_VAR} not found")

    w3 = Web3(Web3.HTTPProvider(eth_rpc))
    if not w3.is_connected():
        print("Failed to connect to Ethereum node.")

    block_range = get_block_range(w3, LOOKBACK_BLOCKS)
    print(f"Fetching blocks from {block_range} range.")
    data = fetch_block_data(block_range, eth_rpc)
    if not data.empty:
        num_entries = len(data)
        print(f"Number of blocks fetched: {num_entries}")
        plot_top_authors(data, num_entries)

if __name__ == "__main__":

Here's a step-by-step breakdown of what this script does:

  1. Setting Up the Environment: We start by importing necessary libraries like os, time, pandas, matplotlib.pyplot, and Web3, along with cryo. Then, we define constants for the RPC URL, the number of blocks to look back on, and the number of top authors to display.
  2. Fetching Blockchain Data: We define a function to determine the range of blocks to fetch based on the current block number. Another function uses cryo.collect to get data on these blocks and returns it as a pandas DataFrame. We track the time taken for this operation, offering insights into the performance of our data retrieval process.
  3. Data Visualization: With the blockchain data in hand, we analyze the top block authors using a function that counts the occurrences of each author in the data. We then use matplotlib to create a bar chart, showcasing the top authors based on the number of blocks mined.
  4. Executing the Script: In the main function, we initialize a Web3 instance, connect to the Ethereum node, fetch the block data, and, if successful, visualize the top authors. We handle potential errors, such as missing environment variables or connection issues, to ensure robustness.
  5. Running the Code: This script is designed as a standalone program. When executed, it will display a bar chart illustrating the most active Ethereum block authors over a specified block range.

This example demonstrates how to effectively combine cryo with other Python tools to fetch, process, and visualize Ethereum blockchain data, providing valuable insights into blockchain activity.

Here is an example of the console output and chart. The console will output something like the following:

Fetching blocks from 18874064:18879064 range.
Time taken to fetch blocks: 30.83 seconds
Number of blocks fetched: 5000

Top Authors by Number of Blocks Mined:
0x95222290dd7278aa3ddd389cc1e1d165cc4bafe5: 1417 blocks
0x1f9090aae28b8a3dceadf281b0f12828e676c326: 1385 blocks
0x4838b106fce9647bdf1e7877bf73ce8b0bad5f97: 443 blocks
0x388c818ca8b9251b393131c08a736a67ccb19297: 295 blocks
0xb9342d6a9789cc6479e48cfef67590c1bd05744e: 213 blocks
0x88c6c46ebf353a52bdbab708c23d0c81daa8134a: 183 blocks
0xdafea492d9c6733ae3d56b7ed1adb60692c98bc5: 175 blocks
0x0aa8ebb6ad5a8e499e550ae2c461197624c6e667: 89 blocks
0x4675c7e5baafbffbca748158becba61ef3b0a263: 55 blocks
0x690b9a9e9aa1c9db991c7721a92d351db4fac990: 52 blocks

And the chart will look like this:

Visualise ERC-20 balance changes over time

The next example weโ€™ll work on will use the same erc20_balances dataset used in one of the previous examples. This time, weโ€™ll fetch and visualize how much WETH is in theWETH-USDT pool from Uniswap V2.

In a new file, paste the following code:

import os
import time
from dotenv import load_dotenv
import cryo
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
from web3 import Web3

# Constants
LOOKBACK_BLOCKS = 7200 # Approx a day in the past
CONTRACT_ADDRESS = '0xC02aaA39b223FE8D0A0e5C4F27eAD9083C756Cc2' # WETH
WALLET_ADDRESS = '0x0d4a11d5EEaaC28EC3F61d100daF4d40471f1852'   # WETH-USDT pool Uniswap V2

# Initialize environment variables and Web3
eth_rpc = os.getenv(ETH_RPC_VAR)
w3 = Web3(Web3.HTTPProvider(eth_rpc))

def check_eth_rpc_connection(eth_rpc):
    """Check connection to Ethereum RPC."""
    if not eth_rpc:
        raise ValueError(f"Environment variable {ETH_RPC_VAR} not found")
    if not w3.is_connected():
        raise ConnectionError("Failed to connect to Ethereum node.")

def get_block_range(lookback_blocks):
    """Determine the range of blocks to fetch."""
    latest_block = w3.eth.block_number
    start_block = max(0, latest_block - lookback_blocks)
    return f"{start_block}:{latest_block}"

def fetch_erc20_balances(block_range):
    """Fetch ERC-20 token balances within a given block range."""
    return cryo.collect(
        requests_per_second=900 # Adapt the RPS to your endpoint

def convert_balance_to_ether(balance_str):
    """Convert balance from Wei to Ether, handling None values."""
    return None if balance_str is None else Web3.from_wei(int(balance_str), 'ether')

def plot_balance_change_over_time(data):
    """Plot the balance change over time on a chart."""
    plt.figure(figsize=(12, 6))
    plt.plot(data['block_number'], data['balance_ether'], marker='o')

    # Set axis labels and chart title with contract and address
    plt.xlabel("Block Number")
    plt.ylabel("Balance (Ether)")
    plt.title(f"ERC-20 Token Balance Change for {CONTRACT_ADDRESS}\nWallet {WALLET_ADDRESS}")

    # Manually set the x-axis ticks based on the block range
    block_numbers = data['block_number']
    tick_spacing = (block_numbers.max() - block_numbers.min()) // 10  # for example, 10 evenly spaced ticks
    ticks = range(int(block_numbers.min()), int(block_numbers.max()), int(tick_spacing))
    plt.xticks(ticks, [f"{tick:,.0f}" for tick in ticks])

    # Format the y-axis to show balances rounded to 4 decimal places

    # Add grid, tighten layout, and display the plot

def main():
    """Main function to fetch and plot ERC-20 token balance changes."""
    block_range = get_block_range(LOOKBACK_BLOCKS)

    # Start timing the data fetch
    start_time = time.time()

    # Fetch the data
    data = fetch_erc20_balances(block_range)

    # End timing the data fetch
    end_time = time.time()
    elapsed_time = end_time - start_time
    print(f"Data fetched in {elapsed_time:.2f} seconds.")

    if data.empty:
        print("No data available for plotting.")
    # Prepare data for plotting
    data = data[['block_number', 'erc20', 'address', 'balance_string']]
    data['balance_ether'] = data['balance_string'].apply(convert_balance_to_ether)
    data = data[data['balance_ether'].notnull()]  # Filter out rows with None values

    # Print data summary to the console
    print("\nData summary:")
    print(f"Block rage: {block_range}")
    print(f"Start balance in Ether: {data.iloc[0]['balance_ether']}")
    print(f"End balance in Ether: {data.iloc[-1]['balance_ether']}")

    # Plot the balance changes over time

if __name__ == "__main__":

Here's a step-by-step explanation of whatโ€™s going on:

  1. Fetch Block Range:
    • It calculates the range of blocks to query by finding the latest block number and subtracting the lookback period to determine the start block; in the example, we analyze about a day's worth of blocks.
  2. Fetch ERC-20 Balances:
    • The script fetches ERC-20 token balance data from the specified contract, wallet address, and block range. The cryo.collect function is called, and the data is returned in a pandas DataFrame format.
  3. Data Conversion and Cleaning:
    • A conversion function transforms balance values from Wei (the smallest unit of Ether) to Ether for readability. It handles any None values to avoid errors during conversion.
  4. Summarizing Data:
    • The script prints out a summary of the data to the console, including the block range and the start and end balances in Ether, providing a quick overview of the dataset.
  5. Data Visualization:
    • It then plots the balance changes over time using matplotlib. The x-axis represents block numbers, and the y-axis represents the balance in Ether.
    • The axis tick labels are formatted for better readability, and the chart is titled with the contract and wallet address for reference.


Remember to adapt the request per second.

Here is an example of the result:

Data fetched in 16.51 seconds.

Data summary:
Block rage: 18872292:18879492
Start balance in Ether: 27034.167858289615425314
End balance in Ether: 27039.186876795597977365

Graph for the balance change over a day:

As you can see, we can use cryo to fetch data and manipulate it with Python, a very powerful combo.


The integration of cryo with Python is a significant advancement for blockchain data analysis. It combines cryo's efficient data extraction capabilities with Python's powerful data processing and visualization tools. This synergy, coupled with high-performance Chainstack GENs, enables users to easily extract, analyze, and visualize blockchain data, making it an invaluable resource for developers, researchers, and enthusiasts in the blockchain community. The practical examples demonstrate this integration's real-world utility, highlighting its potential to yield insightful and actionable information from complex blockchain datasets. In essence, cryo and Python offer an effective and accessible platform for in-depth blockchain data exploration.

About the author

Davide Zambiasi

๐Ÿฅ‘ Developer Advocate @ Chainstack
๐Ÿ› ๏ธ BUIDLs on EVM, The Graph protocol, and Starknet
๐Ÿ’ป Helping people understand Web3 and blockchain development
Davide Zambiasi | GitHub Davide Zambiasi | Twitter Davide Zambiasi | LinkedIN