Collections
Read each collection’s status, optimizer status, points, indexed vectors, segments, shards and replication factor.
Qdrant connector · Vector database reports
Bring Qdrant collections and cluster status into your reports and AI assistants: each collection’s status, optimizer status, points and indexed vectors, vector settings and quantization, and the cluster’s Raft health.
Updated November 2026
| Collection name | Status | Points count |
|---|---|---|
| invoices | green | 1,284,310 |
| vendors | yellow | 96,412 |
| contracts | green | 18,920 |
A Qdrant cluster, in Qdrant Cloud or self-hosted, is the data source, with all collections or the one you pick. Reports show the cluster as it is when they run.
Read each collection’s status, optimizer status, points, indexed vectors, segments, shards and replication factor.
See vector names, dimension, distance, sparse vectors, payload indexes and quantization.
Check Qdrant version, peers, Raft role, leader, term, pending operations and peers with send failures.
Choose one collection from the list, or all of them, when you add Qdrant.
2 reporting datasets, each with its own fields. Pick one in the app, then choose your columns.
Each Qdrant collection with its health, point and indexed vector counts, segments, vector settings, shards and payload indexes.
Qdrant server version, whether clustering is on, peers, this peer's Raft role, term, commit index and pending operations.
27 report fields across 2 datasets.
Search the available fields. Hover or tap a field to see what it means.
No fields match. Try another search or choose All.
Using MCP, Claude can find Qdrant collections whose indexed vectors trail their points, and check whether the cluster leader changed.
Example
Which collections are not green, and how far behind is their indexing?
List collections whose status isn’t green with optimizer status, and subtract indexed vectors from points to show each backlog.
Datapipe is a hosted MCP server. Add it once as a custom connector in Claude, ChatGPT or Grok, and your assistant can work with Qdrant alongside the other data sources in your project.
https://trydatapipe.com/mcp as a custom connector in your assistant.Qdrant collections with status, optimizer status, points, indexed vectors, vector settings and quantization, plus cluster version, peers and Raft state. Qdrant is read-only in Datapipe.
Your Qdrant cluster URL and a read-only database key from the cluster’s API Keys page. Self-hosted servers need a public https address.
Read-only. Qdrant Cloud keys expire after 90 days by default, and a key limited to some collections can’t read cluster status. Up to 20,000 rows per report.
Create a read-only database key on your Qdrant cluster, then add Qdrant under Data sources.
Enter your cluster URL and key, then pick a collection or all of them.
Send collection status to a sheet each hour, or ask ChatGPT which collections are behind on indexing.
Add both to the same project. Each keeps its own report, so you can check them side by side or ask ChatGPT or Claude a question that needs both.
Collections with status, counts and vector settings, and cluster status with Raft details.
Yes, when the server has a public https address. Qdrant Cloud works too.
A read-only database key. Keys limited to some collections can’t read cluster status.
No. It reads counts, settings and status, never point vectors or payloads.
Yes. Datapipe is a hosted MCP server that covers Qdrant and the other data sources in your project. Add https://trydatapipe.com/mcp as a custom connector in Claude, ChatGPT or Grok, then sign in to Datapipe. Qdrant collections with status, optimizer status, points, indexed vectors, vector settings and quantization, plus cluster version, peers and Raft state. Qdrant is read-only in Datapipe.
Free to start. Paid plans add more data sources and refresh faster.
Free and Starter refresh once a day. Pro refreshes every hour. Agency refreshes every 15 minutes on the first 50 data sources.
One for each Qdrant account you add to a project. Adding the same account to another project counts again. Reports and destinations do not use extra data sources.
Catch a yellow collection or a stuck optimizer before search results start missing.