Text search

Firestore does not support native text search, so we need to rely on external solutions. If you specify a textSearchEnabled flag to the collection, you will see a search bar on top of the collection view.

Search Options

Option	Cost	Setup	Best For
Typesense Extension (Recommended)	~$7-14/month flat	5 min	Most projects
Algolia	Per-query pricing	15 min	Enterprise, geo-search
Local Text Search	Free	1 min	Small collections (<1000 docs)

Using Typesense (Recommended)

The FireCMS Typesense Extension deploys a Typesense search server on a Compute Engine VM and automatically syncs your Firestore data. Features:

🔍 Typo-tolerant search - “headphnes” matches “headphones”
⚡ Sub-millisecond responses
💰 Flat monthly cost - No per-query charges
🔄 Real-time sync - Documents auto-index on create/update/delete

Installation

Prerequisites:

Firebase project with Firestore
GCP billing enabled
gcloud CLI installed

Step 1: Install the extension

firebase ext:install https://github.com/firecmsco/typesense-extension --project=YOUR_PROJECT_ID

Step 2: Grant permissions

export PROJECT_ID=your-project-id
export EXT_INSTANCE_ID=typesense-search  # Default extension name

gcloud projects add-iam-policy-binding $PROJECT_ID \
  --member="serviceAccount:ext-${EXT_INSTANCE_ID}@${PROJECT_ID}.iam.gserviceaccount.com" \
  --role="roles/compute.admin" --condition=None

gcloud projects add-iam-policy-binding $PROJECT_ID \
  --member="serviceAccount:ext-${EXT_INSTANCE_ID}@${PROJECT_ID}.iam.gserviceaccount.com" \
  --role="roles/secretmanager.admin" --condition=None

gcloud projects add-iam-policy-binding $PROJECT_ID \
  --member="serviceAccount:ext-${EXT_INSTANCE_ID}@${PROJECT_ID}.iam.gserviceaccount.com" \
  --role="roles/datastore.user" --condition=None

Step 3: Provision the search server

curl "https://REGION-PROJECT_ID.cloudfunctions.net/ext-typesense-search-provisionSearchNode"

Replace REGION with your functions region (e.g., us-central1) and PROJECT_ID with your project.

Wait ~2 minutes. Existing documents are automatically indexed.

Step 4: (Optional) Enable public search access

gcloud functions add-iam-policy-binding ext-${EXT_INSTANCE_ID}-api \
  --member="allUsers" \
  --role="roles/cloudfunctions.invoker" \
  --region=REGION \
  --project=${PROJECT_ID}

Using Typesense in FireCMS Cloud

Navigate to Project Settings and configure:

Setting	Value
Region	Your extension’s region (e.g., `us-central1`)
Extension Instance ID	Default: `typesense-search`

That’s it! FireCMS Cloud automatically connects to your Typesense instance.

Using Typesense in Self-Hosted FireCMS

import { buildFireCMSSearchController, useFirestoreDelegate } from "@firecms/firebase";

const textSearchControllerBuilder = buildFireCMSSearchController({
  region: "us-central1",  // Your extension's region
  extensionInstanceId: "typesense-search"  // Default name
});

export function App() {
  const firestoreDelegate = useFirestoreDelegate({
    firebaseApp,
    textSearchControllerBuilder
  });
  // ... rest of your app
}

Using Typesense Directly (Without FireCMS)

Search via the API proxy endpoint:

const response = await fetch(
  "https://REGION-PROJECT_ID.cloudfunctions.net/ext-typesense-search-api/collections/products/documents/search",
  {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      q: "blue wireless headphones",
      query_by: "name,description"
    })
  }
);
const results = await response.json();

See Typesense API docs for all available endpoints.

Using Algolia

Algolia is a managed search service with per-query pricing. Best for enterprise needs or advanced features like geo-search.

You need to define a FirestoreTextSearchControllerBuilder and add it to your config. Set up an Algolia account and sync documents using their Firebase extension.

Using Algolia in FireCMS Cloud

We provide a utility method for performing searches in Algolia performAlgoliaTextSearch. You need to import the algoliasearch library and create an Algolia client. Then you can use the performAlgoliaTextSearch method to perform the search.

In your controller, you can define the paths you want to support and the search. The paths not specified can still be searched with the local text search.

Example:

import { algoliasearch, SearchClient } from "algoliasearch";

import {
    performAlgoliaTextSearch,
    buildExternalSearchController,
    FirestoreTextSearchController,
    buildCollection,
    FireCMSCloudApp,
    EntityCollectionsBuilder,
    FireCMSAppConfig
} from "@firecms/cloud";

const client: SearchClient | undefined = algoliasearch("YOUR_ALGOLIA_APP_ID", "YOUR_ALGOLIA_SEARCH_KEY");

const algoliaSearchController = buildExternalSearchController({
    isPathSupported: (path) => path === "products",
    search: async ({
                       path,
                       searchString
                   }) => {
        if (path === "products") {
            return performAlgoliaTextSearch(client, "products", searchString);
        }
        return undefined;
    }
});


const appConfig: FireCMSAppConfig = {
    version: "1",
    textSearchControllerBuilder: algoliaSearchController,
    // ...
}

Using Algolia in self-hosted FireCMS

For self-hosted FireCMS, you need to define a FirestoreTextSearchControllerBuilder.

import { algoliasearch, SearchClient } from "algoliasearch";

import { buildExternalSearchController, performAlgoliaTextSearch } from "@firecms/firebase";

const client: SearchClient | undefined = algoliasearch("YOUR_ALGOLIA_APP_ID", "YOUR_ALGOLIA_SEARCH_KEY");

const algoliaSearchController = buildExternalSearchController({
    isPathSupported: (path) => path === "products",
    search: async ({
                       path,
                       searchString
                   }) => {
        if (path === "products")
            return performAlgoliaTextSearch(client, "products", searchString);
        return undefined;
    }
});


export function App() {

    // ...
    const firestoreDelegate = useFirestoreDelegate({
        firebaseApp,
        textSearchControllerBuilder: algoliaSearchControllerBuilder
    });
    // ...
}

Local text search

Since FireCMS v3 we provide a local text search implementation. This is useful for small collections or when you want to provide a quick way to search through your data.

However, for larger collections, you will want to use an external search provider, such as Algolia. This is the recommended approach.

You can use local text search in FireCMS Cloud, or in self-hosted versions.

For FireCMS Cloud, you just need to enable it in the UI.

For self-hosted versions, you can enable it by setting the localTextSearchEnabled in useFirestoreDelegate. Then you need to mark each collection with textSearchEnabled: true.

If you have declared an external indexing provider, the local text search will be effective only for the paths not supported by the external provider.

Using an external search provider

When using an external search provider, you need to implement a FirestoreTextSearchController.