← All articles

What to use instead of OpenZeppelin Defender, Chainlink Automation, or Gelato Web3 Functions

Defender has retired, Chainlink Automation is deprecated, and Gelato no longer lists Web3 Functions. Here are four options that are live today, what each one asks you to trust, and the questions to answer before you move your jobs.

Lit Protocol

If you ran jobs on OpenZeppelin Defender, Chainlink Automation, or Gelato Web3 Functions, four options that are live today are Chainlink CRE, OpenZeppelin's open-source Relayer and Monitor, KeeperHub, and Lit Triggers with Lit Actions. They differ most in where your code runs, who holds the signing key, and who can change or stop a job. Chainlink CRE runs your code on a Chainlink network and delivers results through a contract you deploy. OpenZeppelin's tools run on your own servers with a key you choose. KeeperHub runs workflows on its hosted service with a wallet it provisions for you. Lit runs your code inside a secure enclave, where the wallet's signing key also stays.

I wrote this for teams that need to move scheduled jobs, webhook handlers, and event watchers somewhere that will keep running.

What happened to each tool

OpenZeppelin closed Defender sign-ups on June 30, 2025, and retired the platform on July 1, 2026. OpenZeppelin now points users to its open-source Relayer and Monitor, which you host yourself or pay OpenZeppelin to host. In October 2026 it also published a guide for moving Defender workflows to Chainlink CRE.

Chainlink deprecated Automation v1.x on June 30, 2026, and v2.1 on July 31, 2026. Chainlink's docs tell users to move their upkeeps to CRE to avoid a disruption in service.

Gelato's website no longer lists Web3 Functions, and its @gelatonetwork/automate-sdk package on npm is marked as deprecated and no longer maintained. Gelato still sells its relayer, bundler, and rollup products.

The options today

Chainlink CRE runs workflows compiled to WebAssembly on a network of independent Chainlink nodes. Workflows can start from a cron schedule, an HTTP request, or a log on an EVM chain or Solana. The nodes agree on a result and sign a report, and a Chainlink forwarder contract delivers the report to a receiver contract that you deploy. That means you deploy a contract that implements Chainlink's receiver interface, either your own contract or a bridge that forwards calls to it. The workflow ID is a hash of the binary and its configuration, and your receiver can check it. An organization needs deploy access from Chainlink before its first deployment, and Chainlink says its registry quotas are being tuned ahead of general availability.

OpenZeppelin Relayer and Monitor are open-source services you run yourself, under the AGPL-3.0 license. Relayer sends transactions with a signer you configure, such as AWS or Google Cloud KMS, HashiCorp Vault, or Turnkey. Monitor watches chains for events and transactions and sends notifications or calls a script or webhook. Neither tool runs your own jobs on a schedule, so you run the schedule and the job logic on your own infrastructure. OpenZeppelin also offers a managed service that hosts Relayer and Monitor for you. If you host them yourself, you control everything, and you also keep it all running.

KeeperHub is a hosted workflow service that publishes migration guides for teams leaving Defender and Gelato. You build workflows in a visual editor, start them on a schedule, a webhook, an on-chain event, or at block intervals, and write custom logic in a JavaScript code step. Each organization gets a Turnkey wallet, which you can export, and you can put a Safe with an on-chain allowlist in front of it. KeeperHub supports about 20 EVM chains plus Solana transfers, and it labels itself as beta.

Lit Triggers with Lit Actions runs a Lit Action on a cron schedule, when a webhook arrives, or when an event is emitted on Ethereum, Base, Arbitrum, BNB Chain, or Polygon. A Lit Action is a JavaScript function identified by the hash of its code, and it runs inside a secure enclave in the Lit network. The action signs with a Lit wallet, called a PKP, whose key is created in the enclave and stays there, so your contract sees an ordinary address and needs no changes beyond granting that address its role. In ChainSecured mode, a wallet or Safe you control on Base owns the list of which code can use each PKP. The Triggers service that decides when to fire is a server Lit runs. It stores your usage key encrypted with a key the service holds, so the service can run the code in that usage key's group and nothing else.

Questions to answer before you move

Where does the job's logic run? On CRE it runs on Chainlink's nodes, on OpenZeppelin's tools it runs on your servers, on KeeperHub it runs on KeeperHub's service, and on Lit it runs in an enclave. Ask who could change the code between the time you deploy it and the time it signs.

Who holds the signing key, and can you take it with you? Defender relayer keys could not be exported, so every team that left Defender had to switch to a new address and grant it the roles the old one had. Whatever you choose now, add an admin function to your contracts that can replace the automation address, so the next move is one transaction.

What does your contract see as the sender? CRE delivers through a forwarder to a receiver contract. The other three send from an address you can allowlist directly. If your contracts check the sender, this decides how much contract work the move takes.

Who can change or stop a job? On CRE, a workflow in the on-chain registry can be owned by a multisig. On KeeperHub, changes go through account logins and team roles. On Lit in ChainSecured mode, the list of which code hashes can use each wallet is on-chain, while the trigger settings, such as the schedule and whether it is enabled, are changed through your Triggers login. Lit's docs also say that the operator can pause signing, so plan for how you would recover.

What should happen when a run goes wrong? Chainlink's docs say a reverted report can be sent again. Lit's docs say anyone with the usage key can run an action with a made-up event payload, so an action that acts on an event should read it again from an RPC it trusts. Assume any service can run a job twice. Make your contracts reject work they have already done, and cap what each run can move.

Moving a job to Lit

A Defender Action or Gelato function usually maps to one Lit Action. A cron job becomes a schedule trigger, a webhook becomes a webhook trigger, and a Monitor-driven action becomes a chain-event trigger. The posts on running a scheduled job that signs transactions, turning a webhook into a signed transaction, and triggering a transaction when an event happens on another chain show each one with code.

Write the job's limits into the action, such as the most it can send per run and the addresses it can send to. Then create a Lit wallet, grant its address the role your old relayer had, and remove the old address once the new job has run cleanly for a while.

Get started

To create an account, go to the Lit dashboard and click Create an account. Lit Triggers is documented at developer.litprotocol.com/lit-triggers.