Understanding on-chain data gives you an edge if you're evaluating investment opportunities, looking for a job or building your own crypto product.

In this series, I will teach you to use Dune and ChatGPT to become a crypto data wizard 🧙✨. No programming experience required.


  • At the bottom of this guide, there is a link to a video version for those who prefer visual learning 🙂
  • I’m using the free version of ChatGPT, so the prompts in this guide will work for everyone

What data is available on Dune?

Dune has data from Ethereum, Arbitrum, Optimism, BSC, Bitcoin, Solana, Avalanche, Fantom and Ethereum’s Goerli Testnet.

There are 4 categories of data tables on Dune

  • Raw data: Unedited, raw blockchain data
  • Decoded data: Contract calls and events grouped by smart contract.
  • Spellbook: Aggregated tables that are easy to work with. Maintained by benevolent Dune wizards.
  • Community: Datasets produced by third-party teams that bring off-chain data to Dune.

We’re going to start our journey by digging into Spellbook tables. Spells are great because they abstract away the complexity of interpreting blockchain data. There are Spell tables that aggregate DEX trades, Airdrop claims, bridge transfers and more.

Writing Dune queries with ChatGPT

In this lesson, we're going to analyse some NFT data.

Pick an NFT collection you want to investigate. I’m going to use Azuki (random choice, I don't own any).

On the collection’s OpenSea page, click the link to its Etherscan page and copy the contract address.

Open Etherscan link on OpenSea
Copy contract address on Etherscan

We are going to use Dune Spells and ChatGPT to answer these questions:

  • What is the monthly trading volume of this collection?
  • How many NFTs of this collection get sold every month?

This post is for paying subscribers only

Already have an account? Sign in.

GPT Crypto Data Guide #1