The problem TokenGPT solves

Crypto markets are wild, and guessing token prices is like tossing a coin. Traditional tools aren’t built for capturing the complex future of digital assets, leaving users lost and high-risk.

TokenGPT flips the script - just pop in a crypto question, and it simulates future price points, tokenomics scores, and key insights in seconds - all on top verified blockchain data and proven simulation engine. Say goodbye to guesswork and hello to data-driven strategies for navigating tomorrow’s markets!

We want to apply for following categories:

1. Best Overall Project
2. Best use of Marketplace Addon - as we are using Gas Estimation API - Based on current transaction cost estimates, transaction costs are assumed in the simulation.
3. Odos - Odos Token Pricing API: We retrieve precise token pricing, used when liquidity pools for the token are found (via another API - DexScreener it's not an add-on in QuickNode).
4. Coinpaprika - CoinPaprika Crypto Price API: We retrieve token information (name, symbol, logo, links, contract addresses, supply, price).

Challenges we ran into

Odos Token Pricing API: Lacks an endpoint/method for liquidity pool information, requiring the use of two different APIs.

Gas Estimation API: The add-on supports only 2 networks. While the API theoretically supports more, the add-on only covers Ethereum and Polygon

Solana Priority Fee API: The add-on did not work ("method not found"). Tested with different QuickNode endpoints

Blockscout Data Rich Multichain REST API: Requires a separate QuickNode endpoint for each network. The multi-chain endpoint option does not solve the issue. Intended for retrieving the number of tokens in liquidity pool contracts.

CoinApi Crypto Market Data API: Could provide token price and current order book on CEX. However, only a partial order book is accessible (QuickNode does not support the GetCurrentOrderBooksDepth method), and the API is highly unstable, failing about 50% of the time when querying the same request.

QuickNode functions: only work with Node.js. and we had to use Python so we stayed with AWS Lambda