Simple On-chain NFT Metadata Gas Optimization
How to optimize on-chain NFT metadata in a single simple step.
I have recently read the post by Parham Alimardanian where there is a mention of generating metadata for NFTs on the chain, his conclusion was that “You can see that the on-chain method gas usage is a lot higher than the off-chain method.” while it is true, his results were exaggerated by bad use of on-chain Metadata.
In solidity there are 2 costly things, saving data, and computation, but in solidity, there are also free things, such as static calls which do not require any gas usage — moving the computation side of things to a statically called function can have a great impact on gas optimization.
This was the contract Parham used to showcase his results:
There are two issues with this approach:
- The “mintOnChain” function is calling “makeURI”, it calculates the data URL using base64 on the same computation space where the transaction takes place, and this increases the gas price.
- The second issue is the use of Open Zepplin's “ERC721URIStorage” which is needed only for static URIs, it serves no purpose when we define on-chain metadata.
Mitigating these issues, we can create a more optimized contract:
How does it translate to gas savings? by simply moving the calculation of the base64 from a place where it would cost gas, to a place where it can be called statically, and removing extra functionality that is only needed for static URIs, we can reduce almost all the gas overhead from saving on-chain metadata.
As you can see, the contract that is optimized for on-chain metadata costs almost the same amount of gas as the normal off-chain NFT, yet it allows us to create a dynamic NFT.
Conclusion
When writing smart contracts, we sometimes have to think of optimizations that are not valid for normal, web2 operations, doing this change on a normal backend written in Go or Node would change nothing, and would probably increase the overhead since it will not allow caching — but with smart contracts, that is a completely valid way to achieve gas optimization.
