Functions

Functions are serverless compute you deploy onto a branch, so your backend code runs next to your database. A function doesn't configure the services it uses: enable a service on the branch and its connection strings and credentials are injected into process.env at runtime. A function has 15 minutes to start responding and can keep streaming while data flows, which suits agents and WebSocket or SSE servers. Each branch runs its own functions at their own URLs.

The slug is assigned at first deploy and can't be changed; name is a display label only.

Renaming acts only on a function the branch owns. A slug that's only inherited from an ancestor returns 404, so rename it on the owning branch.

You can also manage functions from the CLI with neon functions.

Functions are currently available in AWS US East (Ohio), US East (N. Virginia), Europe (Frankfurt), and Asia Pacific (Singapore). Support is expanding toward all regions. See Neon Functions for the deployment workflow, Environment variables for what gets injected, and Runtime limits for timeouts and concurrency.

API Reference / Functions / Create a trigger

Creates a trigger for a Function visible on the branch. The required type discriminator selects the trigger-specific configuration. The supported types are schedule and storage_object_created. A schedule trigger uses a numeric five-field cron expression interpreted in UTC. A storage-object-created trigger fires only after a successful upload to one exact bucket and may narrow matches to an object-key prefix.

The name must be unique among triggers visible on the branch, including inherited triggers.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID

• type (string, required) Trigger type discriminator. Possible values: schedule
• function_slug (string, required) The branch-local Function slug to invoke.
• name (string, required) Human-readable name, unique among triggers visible on the branch.
• function_path (string, optional) Path passed to the target Function. Defaults to /. Default: /
• schedule (object, required) A numeric five-field cron schedule interpreted in UTC.
  • cron (string, required) Numeric five-field cron expression (minute through day-of-week), interpreted in UTC.
• enabled (boolean, optional) Whether future occurrences should be scheduled. Default: true

• trigger (object, optional) A branch-effective trigger discriminated by type. The supported trigger types are schedule and storage_object_created.

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/triggers" \
  -X POST \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.createProjectBranchTrigger({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Delete a custom domain from a branch

Removes a custom domain registered on the branch and stops routing it.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• domain (string, path, required) The registered custom domain (case-insensitive).

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/custom-domains/$DOMAIN" \
  -X DELETE \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.deleteProjectBranchCustomDomain({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    domain: process.env.DOMAIN
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Delete a function on the branch

Deletes the function identified by its slug.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• slug (string, path, required) The function slug

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/functions/$SLUG" \
  -X DELETE \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.deleteProjectBranchFunction({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    slug: process.env.SLUG
  }
});

# neonctl
neon functions delete <branch_id>

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Delete a trigger

Deletes a branch-local trigger or writes a branch-local tombstone for an inherited trigger so it does not reappear. Deletion stops future scheduling or storage-event matching but does not cancel invocations already committed for delivery. The supported trigger types are schedule and storage_object_created.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• trigger_id (string, path, required) The opaque, project-wide trigger ID

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/triggers/$TRIGGER_ID" \
  -X DELETE \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.deleteProjectBranchTrigger({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    trigger_id: process.env.TRIGGER_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Deploy code to a function

Creates a deployment for the function. Supply any subset of zip, environment, and runtime; omitted fields inherit the function's latest version. At least one field must be supplied. The first deployment of a function must include zip. The newest deployment becomes active.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• slug (string, path, required) The function slug

{
  "deployment": {
    "id": 1,
    "status": "pending",
    "memory_mib": 2048,
    "runtime": "nodejs24",
    "created_at": "2025-01-15T10:30:00Z"
  }
}

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/functions/$SLUG/deployments" \
  -X POST \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.createProjectBranchFunctionDeployment({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    slug: process.env.SLUG
  }
});

# neonctl
neon functions deploy <branch_id>

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Get a trigger

Returns the trigger visible on the branch. The supported trigger types are schedule and storage_object_created.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• trigger_id (string, path, required) The opaque, project-wide trigger ID

• trigger (object, optional) A branch-effective trigger discriminated by type. The supported trigger types are schedule and storage_object_created.

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/triggers/$TRIGGER_ID" \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.getProjectBranchTrigger({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    trigger_id: process.env.TRIGGER_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Get function details

Returns the function identified by its slug.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• slug (string, path, required) The function slug

{
  "function": {
    "id": "myfunction",
    "slug": "myfunction",
    "name": "myfunction",
    "invocation_url": "https://br-young-forest-a5b6c7d8-myfunction.compute.c-5.us-east-2.aws.neon.tech/",
    "current_deployment": {
      "id": 7,
      "status": "completed",
      "memory_mib": 2048,
      "runtime": "nodejs24",
      "created_at": "2025-01-15T10:30:00Z"
    },
    "active_deployment": {
      "id": 7,
      "status": "completed",
      "memory_mib": 2048,
      "runtime": "nodejs24",
      "created_at": "2025-01-15T10:30:00Z"
    },
    "created_at": "2025-01-15T11:00:00Z"
  }
}

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/functions/$SLUG" \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.getProjectBranchFunction({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    slug: process.env.SLUG
  }
});

# neonctl
neon functions get <branch_id>

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / List functions on the branch

Lists functions on the specified branch.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• cursor (string, query, optional) A cursor to use in pagination. A cursor defines your place in the data list. Include response.pagination.next in subsequent API calls to fetch next page of the list.
• limit (integer, query, optional) Specify a value from 1 to 1000 to limit number of functions in the response

{
  "functions": [
    {
      "id": "myfunction",
      "slug": "myfunction",
      "name": "myfunction",
      "invocation_url": "https://br-young-forest-a5b6c7d8-myfunction.compute.c-5.us-east-2.aws.neon.tech/",
      "current_deployment": {
        "id": 4,
        "status": "completed",
        "memory_mib": 2048,
        "runtime": "nodejs24",
        "created_at": "2025-01-15T10:30:00Z"
      },
      "active_deployment": {
        "id": 4,
        "status": "completed",
        "memory_mib": 2048,
        "runtime": "nodejs24",
        "created_at": "2025-01-15T10:30:00Z"
      },
      "created_at": "2025-01-15T11:00:00Z"
    }
  ],
  "pagination": {}
}

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/functions" \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.listProjectBranchFunctions({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID
  }
});

# neonctl
neon functions list

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / List the custom domains on a branch

Lists all custom domains registered on the branch, across every target entity.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• cursor (string, query, optional) A cursor to use in pagination. A cursor defines your place in the data list. Include response.pagination.next in subsequent API calls to fetch next page of the list.
• limit (integer, query, optional) Specify a value from 1 to 1000 to limit number of domains in the response

• custom_domains (array, optional)

  • domain (string, required) The registered custom domain (normalized, lowercase).

  • entity_type (string, required) The kind of branch entity the domain targets. Possible values: function (v1 supports only function). Not an enum: new values may ship in later spec versions — treat any undocumented value as unknown.

  • entity_id (string, required) The target entity's identifier within the branch. For function this is the function slug.

  • cname_target (string, required) The hostname the customer must point their custom domain at with a CNAME record. Empty when the serving region has no custom-domains front door configured. This is the activation input: point DNS here and the domain goes live (see status) once a certificate is issued on the first request.

  • status (string, optional) The domain's current validity, computed by a background check: pending (still converging — point your CNAME at cname_target and wait), active (live: DNS resolves to the edge, the CA is authorized, and routing is published), or error (a fixable problem — see status_reason). Not an enum: treat any undocumented value as unknown. May be absent briefly right after registration.

  • dns_status (string, optional) The DNS + CAA portion of the check: pending (no records yet), ok (resolves to our edge and the CA is authorized), misconfigured (your CNAME does not resolve to our edge), or caa_blocked (your CAA records forbid Let's Encrypt). Not an enum.

  • binding_status (string, optional) Whether Neon's internal routing for the domain is published: pending, present, or missing. missing is an internal fault surfaced for support. Not an enum.

  • status_reason (string, optional) A short, stable machine-readable reason for a non-active status (e.g. cname-not-pointing-at-edge, caa-blocks-lets-encrypt, binding-missing), suitable for keying an actionable hint. Empty when active or pending.

• pagination (object, optional) To paginate the response, issue an initial request with limit value. Then, add the value returned in the response .pagination.next attribute into the request under the cursor query parameter to the subsequent request to retrieve next page in pagination. The contents on cursor next are opaque, clients are not expected to make any assumptions on the format of the data inside the cursor.

  • next (string, optional) Cursor for the next page of results. Pass it as the cursor query parameter on the next request. Absent on the last page.
  • sort_by (string, optional) Field by which the results were sorted, echoing the request's sort_by parameter.
  • sort_order (string, optional) Sort order active for this page. Pass back as sort_order in the next request to maintain consistent ordering. Valid values are asc and desc.

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/custom-domains" \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.listProjectBranchCustomDomains({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / List triggers on the branch

Lists the complete project-bounded set of triggers visible on the branch, ordered by trigger_id. An inherited trigger keeps its project-wide ID and configuration, and is disabled on the child until explicitly enabled there.

The supported trigger types are schedule and storage_object_created. A storage-object-created trigger watches one exact bucket and fires only after an object upload succeeds.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID

• triggers (array, optional)

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/triggers" \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.listProjectBranchTriggers({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Register a custom domain on a branch

Registers a customer-owned domain (for example dashboard.acme.com) on the branch and points it at a target entity, chosen by entity_type + entity_id. In v1 only entity_type: function is supported (an unsupported type is rejected with 400), where entity_id is the function slug and the function must already exist on the branch (else 404).

The response includes the cname_target the customer must point their domain at with a CNAME record; the domain goes live only once that DNS resolves and a certificate is issued on the first request. A domain already registered to another resource is rejected with 409 and no detail about the owner. Re-registering the same domain for the same entity is idempotent.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID

• domain (string, required) The custom domain to register (for example dashboard.acme.com). Case-insensitive; normalized to lowercase (a trailing root dot is stripped, so the 254-char bound admits a fully-qualified name whose normalized form is 253 chars). Neon-managed and internal hostnames are rejected.

• entity_type (string, required) The kind of branch entity to point the domain at. v1 supports only function; any other value is rejected with invalid_entity_type.

• entity_id (string, required) The target entity's identifier within the branch. For function this is the function slug (which must already exist on the branch).

• domain (string, optional) The registered custom domain (normalized, lowercase).

• entity_type (string, optional) The kind of branch entity the domain targets. Possible values: function (v1 supports only function). Not an enum: new values may ship in later spec versions — treat any undocumented value as unknown.

• entity_id (string, optional) The target entity's identifier within the branch. For function this is the function slug.

• cname_target (string, optional) The hostname the customer must point their custom domain at with a CNAME record. Empty when the serving region has no custom-domains front door configured. This is the activation input: point DNS here and the domain goes live (see status) once a certificate is issued on the first request.

• status (string, optional) The domain's current validity, computed by a background check: pending (still converging — point your CNAME at cname_target and wait), active (live: DNS resolves to the edge, the CA is authorized, and routing is published), or error (a fixable problem — see status_reason). Not an enum: treat any undocumented value as unknown. May be absent briefly right after registration.

• dns_status (string, optional) The DNS + CAA portion of the check: pending (no records yet), ok (resolves to our edge and the CA is authorized), misconfigured (your CNAME does not resolve to our edge), or caa_blocked (your CAA records forbid Let's Encrypt). Not an enum.

• binding_status (string, optional) Whether Neon's internal routing for the domain is published: pending, present, or missing. missing is an internal fault surfaced for support. Not an enum.

• status_reason (string, optional) A short, stable machine-readable reason for a non-active status (e.g. cname-not-pointing-at-edge, caa-blocks-lets-encrypt, binding-missing), suitable for keying an actionable hint. Empty when active or pending.

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/custom-domains" \
  -X POST \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.registerProjectBranchCustomDomain({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Update a function

Updates the function's mutable metadata — currently only the display name. A string sets the display name; null clears it, after which the function's name falls back to its slug. Leading and trailing whitespace is trimmed; a whitespace-only name is rejected. Acts only on a function owned by the branch: a slug that is only inherited from an ancestor branch returns 404 — rename it on the branch that owns it. Like every other change on a branch, a rename is isolated per branch: a branch forked before the rename keeps the name it had at fork time.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• slug (string, path, required) The function slug

• name (string, required) New display name for the function. null clears the display name; the function's name then falls back to its slug. Leading and trailing whitespace is trimmed; a whitespace-only name is rejected.

{
  "function": {
    "id": "myfunction",
    "slug": "myfunction",
    "name": "My Function",
    "invocation_url": "https://br-young-forest-a5b6c7d8-myfunction.compute.c-5.us-east-2.aws.neon.tech/",
    "current_deployment": {
      "id": 10,
      "status": "completed",
      "memory_mib": 2048,
      "runtime": "nodejs24",
      "created_at": "2025-01-15T10:30:00Z"
    },
    "active_deployment": {
      "id": 10,
      "status": "completed",
      "memory_mib": 2048,
      "runtime": "nodejs24",
      "created_at": "2025-01-15T10:30:00Z"
    },
    "created_at": "2025-01-15T11:00:00Z"
  }
}

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/functions/$SLUG" \
  -X PATCH \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.updateProjectBranchFunction({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    slug: process.env.SLUG
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message

API Reference / Functions / Update a trigger

Applies a partial update. The required type discriminator must identify the existing trigger kind. The supported types are schedule and storage_object_created. Editing an inherited trigger creates a child-local shadow with the same trigger_id; it remains disabled unless this request explicitly enables it. For a schedule trigger, updating the schedule or enabled state increments version and recomputes next_run_at.

Disabling stops future scheduling but does not cancel occurrences already committed for delivery. For storage_object_created, the configuration selects one exact bucket. An omitted object-key prefix matches every key in that bucket; a present prefix is matched byte-for-byte and case-sensitively against the full key, without path normalization or a path-segment boundary.

Note: This endpoint is currently in Beta.

• project_id (string, path, required) The Neon project ID
• branch_id (string, path, required) The Neon branch ID
• trigger_id (string, path, required) The opaque, project-wide trigger ID

• type (string, required) Trigger type discriminator; it does not change the trigger type. Possible values: schedule
• function_slug (string, optional) Replacement branch-local Function slug.
• name (string, optional)
• function_path (string, optional)
• schedule (object, optional) A numeric five-field cron schedule interpreted in UTC.
  • cron (string, required) Numeric five-field cron expression (minute through day-of-week), interpreted in UTC.
• enabled (boolean, optional) True enables and false disables future scheduling.

• trigger (object, optional) A branch-effective trigger discriminated by type. The supported trigger types are schedule and storage_object_created.

curl "https://console.neon.tech/api/v2/projects/$PROJECT_ID/branches/$BRANCH_ID/triggers/$TRIGGER_ID" \
  -X PATCH \
  -H "Authorization: Bearer $NEON_API_KEY"

import { createNeonClient, raw } from '@neon/sdk';

const neon = createNeonClient({ apiKey: process.env.NEON_API_KEY });
const { data } = await raw.updateProjectBranchTrigger({
  client: neon.client,
  path: {
    project_id: process.env.PROJECT_ID,
    branch_id: process.env.BRANCH_ID,
    trigger_id: process.env.TRIGGER_ID
  }
});

default General Error.

The request may or may not be safe to retry, depending on the HTTP method, response status code, and whether a response was received.

• If no response is returned from the API, a network error or timeout likely occurred.
• In some cases, the request may have reached the server and been successfully processed, but the response failed to reach the client. As a result, retrying non-idempotent requests can lead to unintended results.

The following HTTP methods are considered non-idempotent: POST, PATCH, DELETE, and PUT. Retrying these methods is generally not safe. The following methods are considered idempotent: GET, HEAD, and OPTIONS. Retrying these methods is safe in the event of a network error or timeout.

Any request that returns a 503 Service Unavailable response is always safe to retry.

Any request that returns a 423 Locked response is safe to retry. 423 Locked indicates that the resource is temporarily locked, for example, due to another operation in progress.

• request_id (string, optional) Unique identifier for the request, useful for debugging. You can set this value manually by including an X-Request-ID header in the request. If not provided, the value will be generated automatically.

• code (string, required) Machine-readable code classifying the error type. See message for a human-readable explanation. Default: ``

• message (string, required) Error message