100+ integrations are ready to connect to your AI.Connect your first MCP

Typesense MCP server for your AI

Search any collection, filter and facet the results, import documents and see which ones failed, change schemas, and shape results with synonyms, curation and presets, on Typesense Cloud or your own server.

PopMCPTypesense

Which collections do I have, and how many documents are in each?

Supported Typesense tools

Works with the collections you already have. It reads each schema for you, so a search does not fail for want of a field list, and it has tools for both current and older Typesense versions.

166 tools in the full catalog: 86 read and 80 write. 35 load by default, and the rest are found with tool search.

typesense_cloud_get_cluster

Read one cluster's configuration, status and hostnames.

typesense_cloud_list_clusters

List the clusters on the Typesense Cloud account with id, status, name, memory, regions and hostnames.

typesense_get_metrics

Read current CPU, memory, disk and network use on the server.

typesense_list_curation_sets

List curation sets, the rules that pin or hide results for chosen queries, with their items.

typesense_list_synonym_sets

List synonym sets with their items.

typesense_upsert_curation_set

Create a curation set, or replace all of its items.

typesense_upsert_synonym_set

Create a synonym set, or replace all of its items.

typesense_create_analytics_rule

Create an analytics rule.

typesense_list_analytics_rules

List analytics rules, which turn searches and clicks into popular queries, no-hit queries, counters and logs.

typesense_delete_document

Permanently delete one document by its id.

typesense_delete_documents_by_filter

Permanently delete every document matching query filter_by, such as status:=cancelled.

typesense_export_documents

Read documents straight out of a collection, with no ranking, optionally filtered.

typesense_get_document

Read one document by its id.

typesense_import_documents

Add or change many documents in one call.

typesense_index_document

Add one document, or change it with action=upsert (replaces the whole document), update (changes the given fields, must exist) or emplace (changes the given fields, creates when missing).

typesense_update_document

Change some fields of one document and leave the others alone.

typesense_update_documents_by_filter

Change the same fields on every document matching a filter.

typesense_create_collection

Create a collection.

typesense_delete_collection

Permanently delete a collection and all its documents.

typesense_describe_collection

Everything needed to search one collection: its schema, which fields work in query_by, facet_by and sort_by, and a few sample documents.

typesense_get_collection

Read one collection's schema: its fields with types and flags, num_documents, default_sorting_field and synonym_sets.

typesense_list_collections

List the collections on this Typesense server with name, num_documents and created_at.

typesense_overview

Start here.

typesense_update_collection

Change a collection's schema without reindexing it yourself.

typesense_create_api_key

Create an API key limited to named actions and collections, such as a search-only key for a website.

typesense_generate_scoped_search_key

Make a scoped search key from a search-only key, with a filter or expiry users cannot change, such as one customer's documents only.

typesense_list_api_keys

List API keys with id, description, actions, collections, expires_at and the first characters of the value.

typesense_list_aliases

List aliases: each gives name and collection_name.

typesense_list_presets

List search presets: saved sets of search parameters that a search names with its preset parameter.

typesense_list_stopword_sets

List stopword sets.

typesense_upsert_alias

Create an alias, or point an existing one at another collection.

typesense_upsert_preset

Create or replace a search preset.

typesense_upsert_stopword_set

Create or replace a stopword set: words a search should ignore.

typesense_multi_search

Run several searches in one call, across one or more collections, and get a result for each in results.

typesense_search_documents

Search one collection.

typesense_cloud_cancel_cloning_operation

Cancel a clone that has not started.

typesense_cloud_cancel_configuration_change

Cancel a configuration change that has not started.

typesense_cloud_cancel_restart

Cancel a rolling restart that has not started.

typesense_cloud_create_cloning_operation

Copy a cluster and its data into a new cluster, for example for staging.

typesense_cloud_create_cluster

Create a cluster on Typesense Cloud.

typesense_cloud_create_configuration_change

Schedule a change to a cluster's size, version, availability or regions.

typesense_cloud_create_restart

Restart Typesense on every node, one at a time, which applies changed server parameters.

typesense_cloud_create_server_parameter

Set a Typesense server parameter on a cluster.

typesense_cloud_delete_server_parameter

Remove a server parameter, going back to its default after the next restart.

typesense_cloud_get_cloning_operation

Read one clone and how far along it is.

typesense_cloud_get_configuration_change

Read one configuration change and how far along it is.

typesense_cloud_get_restart

Read one rolling restart.

typesense_cloud_get_server_parameter

Read one server parameter.

typesense_cloud_list_cloning_operations

List a cluster's clones.

typesense_cloud_list_cluster_options

List the regions, memory, vCPU and disk options a cluster can use.

typesense_cloud_list_configuration_changes

List a cluster's configuration changes, past and pending.

typesense_cloud_list_restarts

List a cluster's rolling restarts.

typesense_cloud_list_server_parameters

List the Typesense server parameters set on a cluster.

typesense_cloud_terminate_cluster

Permanently terminate a cluster and all its data.

typesense_cloud_update_cluster

Rename a cluster or turn automatic capacity upgrades on or off.

typesense_cloud_update_server_parameter

Change a server parameter's value.

typesense_clear_search_cache

Empty the search result cache on this node.

typesense_compact_database

Compact the on-disk database on this node to reclaim disk space.

typesense_get_api_stats

Read recent requests per second and latency per endpoint.

typesense_get_cluster_status

Read this node's status within its cluster.

typesense_get_debug_info

Read the server's version and its state.

typesense_get_health

Check the server is up and ready.

typesense_get_health_with_resource_usage

Check health and include resource usage figures from the server.

typesense_get_schema_changes

See collection schema changes still in progress, with how many documents have been altered and validated so far.

typesense_reset_cluster_peers

Emergency recovery for a cluster stuck without a leader: makes this node reset its list of peers.

typesense_set_slow_request_log

Log requests slower than this many milliseconds on the server.

typesense_take_snapshot

Write a backup snapshot of this node to a folder on the server's own disk, for example query snapshot_path=/tmp/typesense-snapshot.

typesense_trigger_leader_election

Start a leader election from this node.

typesense_delete_collection_override

Delete a curation rule (override) from a collection, on Typesense v29 and older.

typesense_delete_collection_synonym

Delete a synonym from a collection, on Typesense v29 and older.

typesense_delete_curation_set

Delete a curation set and all its items.

typesense_delete_curation_set_item

Delete one item from a curation set.

typesense_delete_synonym_set

Delete a synonym set and all its items.

typesense_delete_synonym_set_item

Delete one item from a synonym set.

typesense_get_collection_override

Read one curation rule (override) of a collection on Typesense v29 and older.

typesense_get_collection_synonym

Read one synonym of a collection on Typesense v29 and older.

typesense_get_curation_set

Read one curation set and its items.

typesense_get_curation_set_item

Read one item of a curation set.

typesense_get_synonym_set

Read one synonym set and its items.

typesense_get_synonym_set_item

Read one item of a synonym set.

typesense_list_collection_overrides

List a collection's curation rules (overrides) on Typesense v29 and older.

typesense_list_collection_synonyms

List a collection's synonyms on Typesense v29 and older.

typesense_list_curation_set_items

List the items in one curation set.

typesense_list_synonym_set_items

List the items in one synonym set.

typesense_upsert_collection_override

Add or replace a curation rule (override) on a collection, on Typesense v29 and older.

typesense_upsert_collection_synonym

Add or replace a synonym on a collection, on Typesense v29 and older.

typesense_upsert_curation_set_item

Add or replace one item in a curation set.

typesense_upsert_synonym_set_item

Add or replace one item in a synonym set.

typesense_cloud_accept_synonym_suggestion

Accept a pending synonym suggestion, which changes live search results.

typesense_cloud_delete_analytics_notification

Stop emailed reports for an analytics rule.

typesense_cloud_delete_synonym_suggestion_settings

Stop synonym suggestions and delete the stored key.

typesense_cloud_dismiss_synonym_suggestion

Dismiss a pending synonym suggestion.

typesense_cloud_get_analytics_notification

Read the report schedule of one analytics rule.

typesense_cloud_get_synonym_suggestion_settings

Read a cluster's weekly synonym suggestion settings.

typesense_cloud_list_analytics_notifications

List the analytics rules that have emailed reports scheduled.

typesense_cloud_list_synonym_suggestions

List synonym suggestions.

typesense_cloud_run_synonym_suggestions

Run synonym suggestions now.

typesense_cloud_set_analytics_notification

Schedule or replace emailed reports for an analytics rule.

typesense_cloud_set_synonym_suggestion_settings

Set up or replace weekly synonym suggestions.

typesense_create_rate_limit_rule

Add a rate limit rule.

typesense_delete_active_throttle

Lift a current limit or ban by its id from typesense_list_active_throttles.

typesense_delete_analytics_rule

Delete an analytics rule.

typesense_delete_rate_limit_exceed

Clear one client's count of going over its limit.

typesense_delete_rate_limit_rule

Delete a rate limit rule.

typesense_flush_analytics

Write the analytics held in memory to disk now, instead of waiting for the next scheduled flush.

typesense_get_analytics_rule

Read one analytics rule.

typesense_get_analytics_status

See how many analytics events are waiting to be written.

typesense_get_rate_limit_rule

Read one rate limit rule by its numeric id.

typesense_list_active_throttles

List the clients being limited or banned right now.

typesense_list_analytics_events

Read one user's recent analytics events for a rule.

typesense_list_rate_limit_exceeds

See how many times each client went over its limit.

typesense_list_rate_limit_rules

List the rate limit rules set on this server.

typesense_send_analytics_event

Record a click, conversion or visit against an analytics rule.

typesense_update_rate_limit_rule

Replace a rate limit rule, with the same body as typesense_create_rate_limit_rule.

typesense_upsert_analytics_rule

Update an analytics rule, or create it under this name.

typesense_cloud_get_access_logs

Read each request's time, method, path, client IP and API key prefix.

typesense_cloud_get_api_keys_last_used

See when each API key was last used, to find ones safe to revoke.

typesense_cloud_get_cluster_metrics

Read per-node metrics for a cluster over a time window, such as memory, CPU and search rate.

typesense_cloud_get_search_insights

Read a collection's top no-hit and popular queries, top facet values and field names.

typesense_cloud_get_slow_requests

Read slow requests, including their search parameters.

typesense_cloud_get_top_api_keys

See which API keys make the most requests.

typesense_cloud_get_top_ips

See which client IPs make the most requests.

typesense_cloud_get_top_paths

See which request paths are used most.

typesense_cloud_get_billing_summary

Read prepaid credit, spend by month, the upcoming invoice and the billing cycle.

typesense_cloud_get_invoice

Read one invoice with its line items, amounts in cents, and invoice_html_url to view it.

typesense_cloud_list_account_usage

List the invoiced node-hours and bandwidth behind the account's invoices.

typesense_cloud_list_cluster_costs

See what each cluster cost per calendar month.

typesense_cloud_list_invoices

List invoices, with amounts in cents.

typesense_cloud_list_prices

List node-hour and bandwidth prices by region and configuration.

typesense_cloud_update_billing_cycle

Switch the account between weekly and monthly billing.

typesense_cloud_create_cluster_api_key

Create a cluster API key limited to named actions and collections.

typesense_cloud_delete_backup

Permanently delete a backup.

typesense_cloud_delete_cluster_api_key

Revoke a cluster API key by its id from typesense_cloud_list_cluster_api_keys.

typesense_cloud_get_backup_settings

Read a cluster's backup settings and their estimated cost.

typesense_cloud_get_ip_allowlist

Read which IPv4 addresses may reach a cluster.

typesense_cloud_list_backups

List a cluster's kept backups.

typesense_cloud_list_cluster_api_keys

List a cluster's API keys with id, description, scope and the first characters of the value.

typesense_cloud_restore_backup

Restore a backup, either onto a new cluster, which is billed, or onto the backed-up cluster, which replaces all its data.

typesense_cloud_set_backup_settings

Turn backups on or off, or change their schedule, retention and regions.

typesense_cloud_set_ip_allowlist

Replace a cluster's whole IP allowlist.

typesense_create_conversation_model

Add a conversation model so a search can answer in sentences with conversation=true.

typesense_create_nl_search_model

Add a natural language search model.

typesense_delete_api_key

Revoke an API key by its numeric id from typesense_list_api_keys.

typesense_delete_conversation_model

Delete a conversation model.

typesense_delete_nl_search_model

Delete a natural language search model.

typesense_get_api_key

Read one API key's details by its numeric id from typesense_list_api_keys.

typesense_get_conversation_model

Read one conversation model.

typesense_get_nl_search_model

Read one natural language search model.

typesense_list_conversation_models

List conversation models, the language model settings used for answers from search results.

typesense_list_nl_search_models

List natural language search models, which turn a sentence into filters and sorts.

typesense_update_conversation_model

Change a conversation model.

typesense_update_nl_search_model

Change a natural language search model.

typesense_delete_alias

Delete an alias.

typesense_delete_preset

Delete a search preset.

typesense_delete_stemming_dictionary

Delete a custom stemming dictionary.

typesense_delete_stopword_set

Delete a stopword set.

typesense_get_alias

See which collection an alias points at.

typesense_get_preset

Read one search preset.

typesense_get_stemming_dictionary

Read a stemming dictionary's word and root pairs.

typesense_get_stopword_set

Read one stopword set.

typesense_import_stemming_dictionary

Create or fill a custom stemming dictionary from word and root pairs.

typesense_list_stemming_dictionaries

List custom stemming dictionaries by id.

typesense_delete_personalization_model

Delete a personalization model.

typesense_get_personalization_model

Read one personalization model.

typesense_list_personalization_models

List uploaded personalization models.

typesense_cloud_get_account

Which Typesense Cloud account the Cluster Management API key belongs to, and what it may do (capabilities).

Connect Typesense to Claude, ChatGPT and Claude Code

Credentials off one Typesense screen, then an endpoint your client signs into. Nothing to deploy, and nothing to install.

On Typesense Cloud, open your cluster and press Generate API Keys to get its hostname and keys. Paste the hostname and the admin key here. On your own server, use its https address and the key it was started with. A Cluster Management API key is optional and adds clusters, backups and billing.

Server addressabc123.a1.typesense.net
API key••••••••••••3f7a
Cluster Management API key••••••••••••3f7a
ConnectionVerified

Then just ask for the work

  • “Which collections do I have, and how many documents are in each?”

    • overview
    • list_collections
  • “Search products for trail shoes under 100, sorted by rating, and show how many there are per brand”

    • describe_collection
    • search_documents
  • “Import these products and tell me exactly which ones failed and why”

    • import_documents
    • get_collection
  • “Make sneakers also find trainers, and pin product 422 to the top when someone searches for sneakers”

    • upsert_synonym_set
    • upsert_curation_set
    • update_collection

Use Typesense from the assistant you already open

One hosted endpoint, registered once per client. The same tools and the same permissions, whichever window the request comes from.

Claude DesktopCustom connector

Add the PopMCP URL under Customize → Connectors and the tools appear in every chat.

Claude WebCustom connector

The same endpoint, registered in the browser. Nothing to install on the machine you are working from.

ChatGPTCustom connector

Register PopMCP as a custom connector, then let ChatGPT read and update the account.

CodexDesktop & CLI

Wire the endpoint into Codex or the Codex CLI when a task needs live business context, not a fixture.

OpenCodeCoding agent

Register the remote MCP server in OpenCode's config for terminal-first and scripted agent runs.

ZCodeDesktop app

Add PopMCP as an MCP server in Z.ai's desktop app and work the account from the same window.

Typesense MCP server questions

What teams ask before putting Typesense behind an AI client.

What does the Typesense MCP server need to connect?

Your server's address and an API key. On Typesense Cloud, open your cluster and press Generate API Keys, then paste the hostname, which looks like abc123.a1.typesense.net, and the admin key. On your own server, use its https address and the key it was started with. PopMCP checks that the address is Typesense before it sends the key, and stores the key encrypted.

Does it work with a self-hosted Typesense server?

Yes, when the server can be reached from the internet over https. A custom port and a reverse proxy path both work. A server on plain http, on a private network or only on your own computer cannot be connected, because PopMCP refuses to send a key unprotected or to reach private addresses.

Can I connect a search-only key?

Yes. The AI can then search, but Typesense will refuse anything else, including listing your collections, so name the collection when you ask. If the key is limited to certain collections, name one of them under More options when you connect, because that is the only way to check such a key. You can also set the connection itself to read-only in PopMCP.

Which Typesense versions does it work with?

Current and older ones. Version 30 replaced per-collection synonyms and overrides with synonym sets and curation sets, and the server has tools for both. When one kind answers not found, the error names the tool for the other kind, and typesense_overview shows the version your server runs.

How does it handle large imports and exports?

An import takes up to 10,000 documents in a call and reports exactly which ones failed and why, because Typesense answers with success even when some documents were rejected. An export stops after the number of documents you set, up to 1,000, and says when there are more. A full export of a big collection is better run from your own machine.

What is the optional Cluster Management API key for?

It adds Typesense Cloud's own management tools: clusters, resizing, restarts, backups, logs, metrics, search insights and billing. Create it in your Typesense Cloud account page, pick what it may do, and paste it under More options. You do not need it to search or change data, and self-hosted servers do not use it.

Can the AI see or create API keys?

It can list keys, which shows only their first characters. It can create a key limited to named actions and collections, such as a search-only key for a website, and that answer holds the value once. It cannot create a key with the wildcard action, because that value would land in your chat: make admin keys in Typesense itself. Language model keys stored on Typesense are hidden from the AI too.

How many tools does the Typesense MCP server have?

166. Thirty-five everyday tools load by default and the rest stay searchable, so your AI client is not crowded. They were checked against Typesense's official specifications and its server's own route table, and every endpoint is either covered or left out for a stated reason, such as internal routes or ones that would put an admin key in your chat.

Do I have to run or host the Typesense MCP server myself?

No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.

Which AI clients can use this MCP server?

Any client that supports a remote Streamable HTTP MCP server: Claude Desktop, Claude Web, Claude Code, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.

How do I connect Typesense to Claude Code?

Connect Typesense in PopMCP first, then run claude mcp add --transport http popmcp https://app.popmcp.com/mcp in a terminal. Inside Claude Code, run /mcp and sign in to PopMCP in the browser window it opens, ticking Typesense on the consent screen. The Typesense tools are then available in that session, and no Typesense credential is stored on your machine.

Why do only 35 of the 166 tools load by default?

Loading 166 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 35-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. Every tool stays searchable, and on Scale the owner can load the full catalog for a teammate's connection.

What is the difference between a tool and a trigger?

A tool is something your AI calls, and there are 166 here. A trigger is an event PopMCP watches for on its own. Typesense has no triggers yet, so everything on this page happens when your AI calls a tool.

Connect Typesense to your AI in minutes

Launch a hosted MCP endpoint for Typesense, keep tool access under control, and start testing real workflow prompts right away.

Connect your first MCP