diff --git a/README.md b/README.md
index 66d2a5a..3082731 100644
--- a/README.md
+++ b/README.md
@@ -36,7 +36,7 @@ Also check out the [cmd/rpc_tester](https://github.com/dipdup-io/starknet-indexe
 
 ## Public instances
 Public deployments with reasonable rate limits are available for testing and prototyping:
-- [Starknet mainnet](https://play.dipdup.io/?endpoint=https://starknet-mainnet-gql.dipdup.net/v1/graphql) `https://starknet-mainnet-gql.dipdup.net/v1/graphql`
+- [Starknet mainnet](https://ide.dipdup.io/?resource=mainnet.starknet.dipdup.net/v1/graphql) `https://starknet-mainnet-gql.dipdup.net/v1/graphql`
 
 ## Notes
 - Indexer works on top of the API provided by the sequencer node — it contains the most comprehensive data set, in particular classes and ABIs ordinary nodes do not always have; It's possible though to outsource several request types to the node API to reduce the load on the sequencer and speed up the indexing process, there's an option in the config for that.