Skip to content

Connectors · AI

Qdrant Cloud MCP server for AI agents.

actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Qdrant Cloud and the rest of your systems. It gives your agent 59 Qdrant Cloud actions (27 reads, 19 writes and 13 that wait for a confirmation) and holds the Qdrant Cloud credential itself, so the agent never does. It paces Qdrant Cloud's limits and records every call.

100,000 free calls a month. No card, no email, no password: press Start a free sandbox on the next screen and keep the key it shows you.

claude mcp add --transport http actionlake https://connect.actionlake.com/mcp
the first call
do {  action: "qdrant_cloud.collections.list",  intent: "check which Qdrant Cloud account is connected"}# List collections · read · 1 upstream call
  • 59actionsin one signed file
  • 27readlook, never change
  • 19writecreate and update
  • 13wait for a yesdestructive, confirmed first

Set up

Connect Qdrant Cloud in three steps.

  1. 01

    Add actionlake to Claude.

    Run the command once in Claude Code; its consent page can create a free sandbox: 100,000 calls a month, no card. In the Claude app, add a custom connector with the URL below and paste your key on the consent page. Creating a new sandbox from the Claude app is not yet tested by us.

    https://connect.actionlake.com/mcp
  2. 02

    Connect Qdrant Cloud.

    API key. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead. actionlake stores what you paste write-only, sealed under your tenant's key.

    • api_keyQdrant API key for the cluster: in Qdrant Cloud open the cluster, then API Keys > Create (a database API key; a…
    • hostWhere your Qdrant Cloud lives: your host
    account { op: "connections.requirements", service: "qdrant_cloud" }
    # free: exactly what to paste, and what is found for you
  3. 03

    Make the first call. Check before a write.

    Ask in plain words; Claude picks the action. Before anything changes, check validates the call and prices it, for free.

    do    { action: "qdrant_cloud.collections.list" }# list collections · read · 1 upstream callcheck { action: "qdrant_cloud.shard_keys.create", input: { … } }# free: inputs, permissions, cost and confirmations, nothing sent

Qdrant Cloud at a glance

Sign-in
API key. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead.
You provide
api_key
Setup fields
host
Limits we pace to
600 calls a minute, per connection
API covered
Qdrant REST API master, at your Qdrant Cloud instance's API base URL, which you enter when connecting
Built from
Qdrant Cloud's documentation, read 23 October 2026OpenAPI spec (GitHub)api.qdrant.tech
Service file
Version 1.0.0, signed (Ed25519)
Cached reads
18 of 27 reads, for up to 1 minute; a write clears what it affects
Shared entities
1, so ask can line Qdrant Cloud up with your other systems
Data
Processed and stored in Sydney, Australia. No US region yet

Capabilities

What an AI agent can do in Qdrant Cloud.

59 actions in 21 areas. Every one carries its effect, so a key can allow the reads and hold the rest.

  • 27readlook, never change
  • 19writecreate and update
  • 13destructivewait for a confirmation
  1. Points

    13

    10 read · 2 write · 1 destructive

    Count points · Facet a payload key with a given filter · Query points · Query points in batch

  2. Collections

    6

    3 read · 2 write · 1 destructive

    List collections · Create collection · Check the existence of a collection · Collection info

  3. Collection snapshots

    4

    1 read · 2 write · 1 destructive

    Create collection snapshot · List collection snapshots · Recover from a snapshot · Delete collection snapshot

  4. Point payload

    4

    0 read · 2 write · 2 destructive

    Overwrite payload · Set payload · Clear payload · Delete payload

  5. Shard snapshots

    4

    1 read · 2 write · 1 destructive

    Create shard snapshot · List shards snapshots for a collection · Recover from a snapshot · Delete shard snapshot

  6. Full snapshots

    3

    1 read · 1 write · 1 destructive

    Create storage snapshot · List of storage snapshots · Delete storage snapshot

  7. Shard keys

    3

    1 read · 1 write · 1 destructive

    Create shard key · List shard keys · Delete shard key

  8. Aliases

    2

    1 read · 1 write · 0 destructive

    List collections aliases · Update aliases of the collections

13 more areas · 20 actions
  • Cluster 2
  • Collection cluster 2
  • Issues 2
  • Named vectors 2
  • Payload indexes 2
  • Point vectors 2
  • Quotas 2
  • Cluster peers 1
  • Cluster telemetry 1
  • Collection aliases 1
  • Optimizations 1
  • Server 1
  • Telemetry 1

Top actions

The Qdrant Cloud API, shaped for agents.

The ids your agent types, what each does, and the one name it shares with other systems. 2 of Qdrant Cloud's 2 action names are shared with at least one other connector, so an agent learns them once.

ActionWhat it doesEffectSame name in
qdrant_cloud.collections.listList collectionsList collections.readcontent.table.list21 connectors
qdrant_cloud.server.getReturns information about the running Qdrant instanceReturns information about the running Qdrant instance.readplatform.health.get18 connectors
qdrant_cloud.aliases.listList collections aliasesList collections aliases.readthis connector
qdrant_cloud.points.countCount pointsCount points.readthis connector
qdrant_cloud.points.facetFacet a payload key with a given filterFacet a payload key with a given filter.readthis connector
qdrant_cloud.shard_keys.listList shard keysList shard keys.readthis connector
qdrant_cloud.shard_keys.createCreate shard keyCreate shard key.writethis connector
qdrant_cloud.collections.createCreate collectionCreate collection.writethis connector
qdrant_cloud.full_snapshots.listList of storage snapshotsList of storage snapshots.readthis connector
qdrant_cloud.named_vectors.createCreate named vectorCreate named vector.writethis connector
qdrant_cloud.full_snapshots.createCreate storage snapshotCreate storage snapshot.writethis connector
qdrant_cloud.issues.clearClear issuesClear issues.destructivethis connector

Try it

Prompts to paste into Claude.

Each one names the Qdrant Cloud actions it will call, so you know what it reads and what it changes before you send it.

  1. Readread

    List collections in Qdrant Cloud and tell me what stands out.

    callsqdrant_cloud.collections.list

  2. Writewrite

    Create shard key in Qdrant Cloud. Show me exactly what you will send before you save it.

    callsqdrant_cloud.shard_keys.create

Governance for Qdrant Cloud

Destructive actions wait for a yes.

13 Qdrant Cloud actions are destructive. These are all of them:

  • Clear issuesqdrant_cloud.issues.clear
  • Delete pointsqdrant_cloud.points.delete
  • Delete shard keyqdrant_cloud.shard_keys.delete
  • Delete collectionqdrant_cloud.collections.delete
  • Clear payloadqdrant_cloud.point_payload.clear
  • Remove peer from the clusterqdrant_cloud.cluster_peers.delete
  • Delete named vectorqdrant_cloud.named_vectors.delete
  • Delete payloadqdrant_cloud.point_payload.delete
  • Delete vectorsqdrant_cloud.point_vectors.delete
  • Delete storage snapshotqdrant_cloud.full_snapshots.delete
  • Delete index for field in collectionqdrant_cloud.payload_indexes.delete
  • Delete shard snapshotqdrant_cloud.shard_snapshots.delete
  • Delete collection snapshotqdrant_cloud.collection_snapshots.delete

A destructive action runs only with confirm: true. Under the human policy the agent cannot confirm for itself: the run stops until a named person approves, in the console or from a signed link in an email. The cap on a key is refused before the call is sent, never after.

doqdrant_cloud.points.deleteconfirm_policy: human

Delete points?

intent: the user asked me to delete points

run suspended · waiting for a named person · key qdrant-cloud-agent cannot confirm itself
In the console, or from an emailed signed link.
a key for this connector
account {  op: "keys.create", name: "qdrant-cloud-agent",  allowed_actions: ["qdrant_cloud.*"], cap_calls: 2000,  confirm_policy: "human_for_writes",  approval_rules: [    { match: "qdrant_cloud.issues.clear", approver: "[email protected]" },    { match: "qdrant_cloud.points.delete", approver: "[email protected]" },    { match: "qdrant_cloud.shard_keys.delete", approver: "[email protected]" }  ]}
Illustrative. Scoped to Qdrant Cloud only, capped at 2,000 upstream calls, every write confirmed, and the destructive ones approved by a named person.

Cost

What Qdrant Cloud costs on actionlake.

Every Qdrant Cloud action is one upstream call, the only thing actionlake bills. Searching, check, cost, account and any call Qdrant Cloud refused with 401, 403, 429 or a 5xx are free. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.

  • 1upstream call per actionall 59 Qdrant Cloud actions
  • 59actions pageeach page of a long list is one call
  • 18reads can be cacheda cache hit costs nothing
  • 100,000calls a month, freepaid plans from US$49 a month for 250,000

See pricing

Action index

Every Qdrant Cloud action.

59 actions by area: the id your agent calls, what it does, and its effect.

Points13
  • qdrant_cloud.points.batch_updateBatch update pointswrite
  • qdrant_cloud.points.countCount pointsread
  • qdrant_cloud.points.deleteDelete pointsdestructive
  • qdrant_cloud.points.facetFacet a payload key with a given filterread
  • qdrant_cloud.points.getGet pointread
  • qdrant_cloud.points.queryQuery pointsread
  • qdrant_cloud.points.query_batchQuery points in batchread
  • qdrant_cloud.points.query_groupsQuery points, grouped by a given payload fieldread
  • qdrant_cloud.points.retrieveGet pointsread
  • qdrant_cloud.points.scrollScroll pointsread
  • qdrant_cloud.points.search_matrix_offsetsSearch points matrix distance offsetsread
  • qdrant_cloud.points.search_matrix_pairsSearch points matrix distance pairsread
  • qdrant_cloud.points.upsertUpsert pointswrite
Collections6
  • qdrant_cloud.collections.createCreate collectionwrite
  • qdrant_cloud.collections.deleteDelete collectiondestructive
  • qdrant_cloud.collections.existsCheck the existence of a collectionread
  • qdrant_cloud.collections.getCollection inforead
  • qdrant_cloud.collections.listList collectionsread
  • qdrant_cloud.collections.updateUpdate collection parameterswrite
Collection snapshots4
  • qdrant_cloud.collection_snapshots.createCreate collection snapshotwrite
  • qdrant_cloud.collection_snapshots.deleteDelete collection snapshotdestructive
  • qdrant_cloud.collection_snapshots.listList collection snapshotsread
  • qdrant_cloud.collection_snapshots.recoverRecover from a snapshotwrite
Point payload4
  • qdrant_cloud.point_payload.clearClear payloaddestructive
  • qdrant_cloud.point_payload.deleteDelete payloaddestructive
  • qdrant_cloud.point_payload.overwriteOverwrite payloadwrite
  • qdrant_cloud.point_payload.setSet payloadwrite
Shard snapshots4
  • qdrant_cloud.shard_snapshots.createCreate shard snapshotwrite
  • qdrant_cloud.shard_snapshots.deleteDelete shard snapshotdestructive
  • qdrant_cloud.shard_snapshots.listList shards snapshots for a collectionread
  • qdrant_cloud.shard_snapshots.recoverRecover from a snapshotwrite
Full snapshots3
  • qdrant_cloud.full_snapshots.createCreate storage snapshotwrite
  • qdrant_cloud.full_snapshots.deleteDelete storage snapshotdestructive
  • qdrant_cloud.full_snapshots.listList of storage snapshotsread
Shard keys3
  • qdrant_cloud.shard_keys.createCreate shard keywrite
  • qdrant_cloud.shard_keys.deleteDelete shard keydestructive
  • qdrant_cloud.shard_keys.listList shard keysread
Aliases2
  • qdrant_cloud.aliases.listList collections aliasesread
  • qdrant_cloud.aliases.updateUpdate aliases of the collectionswrite
Cluster2
  • qdrant_cloud.cluster.getGet cluster status inforead
  • qdrant_cloud.cluster.recoverTries to recover current peer Raft statewrite
Collection cluster2
  • qdrant_cloud.collection_cluster.getCollection cluster inforead
  • qdrant_cloud.collection_cluster.updateUpdate collection cluster setupwrite
Issues2
  • qdrant_cloud.issues.clearClear issuesdestructive
  • qdrant_cloud.issues.getGet issuesread
Named vectors2
  • qdrant_cloud.named_vectors.createCreate named vectorwrite
  • qdrant_cloud.named_vectors.deleteDelete named vectordestructive
Payload indexes2
  • qdrant_cloud.payload_indexes.createCreate index for field in collectionwrite
  • qdrant_cloud.payload_indexes.deleteDelete index for field in collectiondestructive
Point vectors2
  • qdrant_cloud.point_vectors.deleteDelete vectorsdestructive
  • qdrant_cloud.point_vectors.updateUpdate vectorswrite
Quotas2
  • qdrant_cloud.quotas.getGet global quotasread
  • qdrant_cloud.quotas.updateSet global quotaswrite
Cluster peers1
  • qdrant_cloud.cluster_peers.deleteRemove peer from the clusterdestructive
Cluster telemetry1
  • qdrant_cloud.cluster_telemetry.getCollect cluster telemetry dataread
Collection aliases1
  • qdrant_cloud.collection_aliases.listList aliases for collectionread
Optimizations1
  • qdrant_cloud.optimizations.getGet optimization progressread
Server1
  • qdrant_cloud.server.getReturns information about the running Qdrant instanceread
Telemetry1
  • qdrant_cloud.telemetry.getCollect telemetry dataread

Questions

The Qdrant Cloud MCP server, answered.

Is there a Qdrant Cloud MCP server for Claude?

Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 59 Qdrant Cloud actions across points, collections, collection snapshots and point payload, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Qdrant Cloud.

What do I need to connect Qdrant Cloud?

Qdrant API key for the cluster: in Qdrant Cloud open the cluster, then API Keys > Create (a database API key; a…. Plus your host. The free call account { op: "connections.requirements", service: "qdrant_cloud" } lists exactly what to paste.

Can an AI agent delete or change things in Qdrant Cloud?

Only if its key allows it, and only with an explicit confirmation. 13 Qdrant Cloud actions are destructive, including clear issues, delete points and delete shard key. Each runs only with confirm: true. Under the human confirm policy, or an approval rule on the key, the run waits until a named person approves it.

Will an AI agent use up my Qdrant Cloud API limit?

actionlake paces every call to the limits in the Qdrant Cloud service file: 600 calls a minute per connection. Repeated reads are answered from cache (18 Qdrant Cloud actions can be), which never reaches Qdrant Cloud.

What does a Qdrant Cloud action cost?

One upstream call, the only unit actionlake bills. A list that pages bills one call per page (59 Qdrant Cloud actions page). Cache hits, searches, dry runs and calls Qdrant Cloud refused with 401, 403, 429 or a 5xx are free. The Free plan includes 100,000 calls a month with no card. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.

Which Qdrant Cloud API does it cover?

The Qdrant REST API master at your Qdrant Cloud instance's API base URL, which you enter when connecting. It is built from Qdrant Cloud's documentation as read on 23 October 2026. 59 actions in 21 areas, each one in the signed service file (version 1.0.0).

Does actionlake keep my Qdrant Cloud data, and where?

actionlake processes and stores data in Sydney, Australia; there is no US region yet. Reads may be cached for up to 1 minute, and a write clears the reads it affects. Each run's request and response is kept, sealed, for your plan's trace window: 1 day on Free, 3 days on Starter, 7 days on Pro, 14 days on Team, 30 days on Business and 30 days on Enterprise. Your Qdrant Cloud credential is sealed under your tenant's key, never logged and never returned. Your traffic is not used to train models.

Which AI clients can use it?

Claude Code, tested by us end to end, and the Claude app as a custom connector, tested by us with an existing key. Other MCP clients that support remote servers over Streamable HTTP (MCP revisions 2025-06-18 and 2025-11-25) and OAuth dynamic client registration, or a bearer header, should connect the same way; we have not tested them. Or plain REST, under the same action names.

Give your agent Qdrant Cloud today.

Start free - 100,000 calls/month, no card. No signup. Hard caps on every key. Cached answers are free.