Prometheus queries work slow on Grafana 9+

I have some problems with Prometheus on Grafana 9. When I go to explore mode and want to check Prom queries, I get long loading metrics, then I see error ‘Failed to call resource’ and I get ‘No metrics found’, but when I write full query for example kube_pod_info I get all the metrics and graph(it works fine). Any idea why I dont see hints when writing queries? On grafana 8.4.3 everything works fine with same configuration and same prometheus.