Oasis just launched ROFL — Runtime Off-Chain Logic — on mainnet.
This lets you run logic off-chain, outside the blockchain, but still have the results verifiably accepted on-chain. You can think of it like a trustless Lambda function for your smart contracts.
With ROFL, you can:
Fetch external data (APIs, user input)
Run custom business logic off-chain
Perform private AI inference
Return results on-chain with cryptographic proof
It’s built on top of Oasis Sapphire, a confidential EVM that supports TEEs (trusted execution environments). So even though the logic runs outside the blockchain, it’s executed in a secure enclave — and the output comes with a remote attestation proof.
The flow is simple: off-chain logic runs in a TEE, signs the result, and your on-chain contract checks that signature before accepting the result.
This is especially useful for building:
Oracles with custom data sources
Games with private or hidden logic
AI pipelines that integrate onchain
Private voting, auctions, and bridge logic
dApps that can’t afford full onchain compute
You write your logic in Rust or Go, and integrate it easily with Solidity contracts. No special DSL or new language required.
Docs and SDKs are live: https://docs.oasis.io/build/rofl/
If you’ve been looking for a way to break free from onchain compute limits without sacrificing trust assumptions, this might be worth checking out.
Top comments (4)
This is super interesting!
Being able to run logic off chain and still prove the results on chain feels like a game changer, especially for stuff like AI, oracles and anything privacy heavy.
Really appreciate that it works with Rust/Go + Solidity without needing to learn something new. Definitely going to play around with this!
I believe the ROFL framework is the single most important piece of innovation in the decentralized AI space because verifiable privacy with a judicious mix of on-chain and off-chain components is the need of the hour. With the ROFL app now on mainnet, the complexity of building and deploying trustless dApps is also gone. This is a definite game-changer.
For devs who like a practical workshop to get going, this ROFL tutorial will hit the spot.
This is such a clean breakdown of ROFL especially helpful for devs trying to understand how it all fits with TEEs and on-chain verification. I really appreciate how ROFL bridges the gap between Web2-grade computation (like AI workloads) and Web3-level trust. That combo of verifiability + confidentiality is 🔥.
Also love the shoutout to WT3 and Zeph real examples go a long way. Curious to see how far this can scale beyond agents. Maybe confidential DePIN workloads or AI inference marketplaces next?
Solid write-up 👏
Some comments may only be visible to logged-in visitors. Sign in to view all comments.