Everything an AI agent can do with the Postman API.

A reference guide for building AI agents: every method, how to authenticate, and the permissions each one needs.

Endpoints24
API versionv1.39
Last updated23 June 2026
Orientation

How the Postman API works.

The Postman API is how an app or AI agent works with a Postman account: listing and editing collections, managing environments and workspaces, reading API definitions, and running mock servers and monitors. Access is granted through a single API key sent in a request header, and that key inherits the exact permissions of the person who created it, so it reaches whatever Postman data that person can see across their teams. Postman does not push events, but the API can create a webhook whose URL triggers a collection run when it is called.

24Endpoints
8Capability groups
12Read
12Write
0Permissions
Authentication
Every request carries an API key in the X-API-Key header. A key is generated from Postman account settings and provides access to any Postman data the person who created it has permission for. There is no separate token type for machines in the core API, so the key acts as that person across all of their teams and workspaces.
Permissions
Postman has no granular per-endpoint scopes. The API key inherits the role and team memberships of the user who created it, so what a key can do is decided entirely by that user's Postman permissions, not by anything attached to the key itself. A key made by a team admin can do everything that admin can; a key made by a member with read-only access to a workspace cannot write to it. This is why narrowing access has to happen outside Postman, by choosing whose key an agent uses.
Rate limits
Access through an API key is limited to 300 requests per minute overall, with tighter limits on specific endpoints: listing collections, listing and getting workspaces, and the monitors endpoints are capped at 10 calls in 10 seconds, and workspace updates at 20 requests per minute per user. A separate monthly allowance applies per plan. Going over returns HTTP 429.
Data model
The API is resource-oriented JSON over HTTPS at api.getpostman.com. The main resources are collections, environments, workspaces, APIs (the design definitions), mocks (mock servers), and monitors, plus a /me endpoint for the authenticated user. Most resources are addressed by a unique id, and many list endpoints accept a workspace parameter to scope results to one workspace.
Connect & authenticate

Connection & authentication methods.

How an app or AI agent connects to Postman determines what it can reach. The main route is the REST API behind an API key, and Postman also runs a first-party MCP server for agents.

Ways to connect

REST API

The REST API answers at https://api.getpostman.com. Every request carries an API key in the X-API-Key header, and the same base covers collections, environments, workspaces, APIs, mocks, monitors, and the /me endpoint.

Best forConnecting an app or AI agent to Postman.
Governed byThe API key and the permissions of the user who created it.
Docs ↗

MCP server (Model Context Protocol)

Postman's first-party MCP server exposes the Postman API as tools an AI agent can call. The remote server answers at https://mcp.postman.com, with a default minimal tool set at /minimal, a full set of more than 100 tools at /mcp, and a code-focused set at /code, plus an EU host at https://mcp.eu.postman.com. The US server supports OAuth and needs no API key; the EU server uses a Postman API key. The source is at github.com/postmanlabs/postman-mcp-server.

Best forConnecting an app or AI agent to Postman.
Governed byOAuth on the US server, or a Postman API key on the EU server.
Docs ↗

Webhooks

A webhook created through the API returns a URL that, when called, triggers a collection run through a monitor. This is the reverse of an event push: Postman does not notify the app, it gives the app a URL to call when it wants the run to start.

Best forTriggering a collection run on demand from outside Postman.
Governed byPossession of the returned webhook URL.
Docs ↗
Authentication

Postman API key

An API key generated from Postman account settings, sent in the X-API-Key header on every request. The key grants access to any Postman data the person who created it has permission for, so it acts as that user across their teams and workspaces. There are no per-endpoint scopes.

TokenAPI key (X-API-Key header)
Best forConnecting an app or agent to the REST API
Docs ↗

OAuth (MCP server, US)

The US remote MCP server supports OAuth, so an agent connecting through it does not need to hold a long-lived Postman API key. OAuth is not supported on the EU MCP server, which uses an API key instead.

TokenOAuth token (MCP only)
Best forAgents connecting through the US MCP server
Docs ↗
Capability map

What an AI agent can do in Postman.

The Postman API is split into areas an agent can act on, such as collections, environments, workspaces, API definitions, mock servers, and monitors. Each area has its own methods, but all of them are reached through one API key that carries the same access the person behind it has.

Endpoint reference

Every Postman API method.

Filter by method, access, or permission, or search any path. Select a row for version detail, rate limits, the related webhook event, and the source.

MethodEndpointWhat it doesAccessPermissionVersion

Collections

List collections, read a single collection, and create, update, or delete collections.5

No per-endpoint scope exists. The key returns whatever collections its owner can see. This endpoint is rate-limited to 10 calls per 10 seconds.

Acts oncollection
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit10 calls / 10 seconds

No per-endpoint scope exists. Access depends on the key owner's permission on the collection.

Acts oncollection
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The key owner needs write access to the target workspace.

Acts oncollection
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The key owner needs write access to the collection.

Acts oncollection
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The key owner needs delete access to the collection.

Acts oncollection
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

Environments

List environments, read a single environment, and create, update, or delete environments.5

No per-endpoint scope exists. Returns whatever environments the key owner can see.

Acts onenvironment
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. Variable values can include secrets the key owner can read.

Acts onenvironment
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The key owner needs write access to the target workspace.

Acts onenvironment
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The key owner needs write access to the environment.

Acts onenvironment
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The key owner needs delete access to the environment.

Acts onenvironment
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

Workspaces

List workspaces, read a single workspace, and create, update, or delete workspaces.5

No per-endpoint scope exists. Returns every workspace the key owner is a member of. Rate-limited to 10 calls per 10 seconds.

Acts onworkspace
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit10 calls / 10 seconds

No per-endpoint scope exists. Rate-limited to 10 calls per 10 seconds.

Acts onworkspace
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit10 calls / 10 seconds

No per-endpoint scope exists. The key owner needs permission to create workspaces in the team.

Acts onworkspace
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. Workspace-update endpoints are rate-limited to 20 requests per minute per user.

Acts onworkspace
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit20 requests / minute / user

No per-endpoint scope exists. Deleting a workspace removes the collections, environments, and other resources it holds.

Acts onworkspace
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

APIs

List the API definitions in a workspace and read a single API definition.2

No per-endpoint scope exists. The workspaceId query parameter is required to scope the list.

Acts onapi
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. Access depends on the key owner's permission on the API.

Acts onapi
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

Mock servers

List mock servers and create a mock server from a collection.2

No per-endpoint scope exists. Mock servers allow 120 requests per minute when called.

Acts onmock server
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

No per-endpoint scope exists. The created mock server is reachable at the returned mockUrl.

Acts onmock server
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply

Monitors

List monitors, read a single monitor, and run a monitor on demand.3

No per-endpoint scope exists. Monitor endpoints are rate-limited to 10 calls per 10 seconds.

Acts onmonitor
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit10 calls / 10 seconds

No per-endpoint scope exists. Monitor endpoints are rate-limited to 10 calls per 10 seconds.

Acts onmonitor
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit10 calls / 10 seconds

No per-endpoint scope exists. Running a monitor executes its collection against the live systems it targets. Rate-limited to 10 calls per 10 seconds.

Acts onmonitor
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limit10 calls / 10 seconds

Webhooks

Create a webhook whose URL triggers a collection run.1

No per-endpoint scope exists. The request body names the collection to run, and the response returns a webhookUrl that, when called, starts the run.

Acts onwebhook
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventcollection-run-trigger
Rate limitStandard limits apply

User

Read the authenticated user behind the API key.1

No per-endpoint scope exists. Returns the identity and plan usage of whoever created the key.

Acts onuser
Permission (capability)None required
VersionAvailable since the API’s base version
Webhook eventNone
Rate limitStandard limits apply
No endpoints match those filters.
Webhooks

Webhook events.

Postman does not push events to an app or AI agent. Instead, the API can create a webhook that points the other way: a URL Postman gives back that, when called, triggers a collection run through a monitor.

EventWhat it signalsTriggered by
collection run triggerNot an event Postman pushes. A webhook created through the API returns a URL that, when called by anything that holds it, triggers a run of the named collection through a monitor./webhooks
No events match that search.
Rate limits & pagination

Rate limits, pagination & request size.

Postman limits how fast an app or AI agent can call, through an overall per-minute ceiling on each key plus tighter limits on a handful of endpoints and a monthly allowance tied to the plan.

Request rate

Access to the Postman API with a key is limited to 300 requests per minute. Some endpoints are stricter: Postman Monitors and the GET /collections, GET /workspaces, and GET /workspaces/{id} endpoints are limited to 10 calls in any 10-second window, workspace-update endpoints to 20 requests per minute per user, and the service-account-token endpoint to 10 requests per 10-second window per user. Mock servers, public and private, allow 120 requests per minute. A separate monthly usage allowance applies based on the plan. Every response carries X-RateLimit-Limit, X-RateLimit-Remaining, and X-RateLimit-Reset headers, with X-RateLimit-Limit-Month and X-RateLimit-Remaining-Month for the monthly allowance; exceeding a limit returns HTTP 429 with X-RateLimit-RetryAfter and Retry-After giving the seconds to wait.

Pagination

List endpoints page through results with limit and offset query parameters, returning a slice of the full set per call. Many list endpoints also accept a workspace parameter to scope the results to a single workspace rather than every workspace the key can see.

Request size

Requests and responses are JSON. There is no single documented payload size limit across the whole API, though individual resources such as collections and mock servers carry their own plan-based usage limits, like the number of mock-server calls per month.

Errors

Status codes & error handling.

The status codes an agent should handle, and what to do about each.

StatusCodeMeaningWhat to do
401UnauthorizedThe API key is missing, invalid, or expired, so the request was not authenticated.Send a valid key in the X-API-Key header, regenerating it from Postman account settings if needed.
403ForbiddenThe key is valid but the user behind it lacks permission for this resource. Because access follows the user's role, a key cannot exceed what its owner can do.Use a key whose owner has the needed access, or grant that user the permission in Postman.
404Not FoundThe resource id does not exist, or the key's owner cannot see it.Confirm the id and that the key's owner has access to the resource.
429Too Many RequestsA rate limit was exceeded, either the 300-per-minute overall ceiling, a tighter per-endpoint limit, or the monthly allowance.Wait the seconds given in the Retry-After and X-RateLimit-RetryAfter headers, then retry.
Versioning & freshness

Version history.

The Postman API has no version number in its address. It is published as a single collection that Postman revises over time, and the collection carries a version label, currently v1.39, that marks the cut of endpoints.

Version history

What changed, and when

Latest versionv1.39
v1.39Current version
API Catalog and Spec Hub endpoints

The Postman API has no version number in its address; it is published as a public collection that Postman revises, and the collection carries a version label. The v1.39 cut added eight endpoints: six for the API Catalog, covering service monitoring, performance metrics, and compliance data, and two for the Spec Hub, for managing specification version tags and retrieving tagged file snapshots. It was published on 7 May 2026.

What changed
  • Added six API Catalog endpoints for service monitoring, performance metrics, and compliance data
  • Added two Spec Hub endpoints for specification version tags and tagged file snapshots
Earlier
Ongoing additive collection revisions

Earlier collection versions established the core resource endpoints for collections, environments, workspaces, APIs, mocks, monitors, and the authenticated user. New endpoints are added in later versions without breaking existing calls, so the API has no dated breaking-change timeline of the kind a header-versioned API carries.

What changed
  • Established the core collections, environments, workspaces, APIs, mocks, monitors, and user endpoints
  • Added the webhooks endpoint for triggering collection runs

New endpoints arrive in later collection versions without breaking the existing ones.

Postman release notes ↗
Questions

Postman API, answered.

How does an app authenticate to the Postman API?+
By sending an API key in the X-API-Key header on every request. The key is generated from Postman account settings under API keys. There is no OAuth flow for the core REST API; the key alone identifies and authorizes the caller, and it grants access to any Postman data the person who created the key has permission for.
Can I limit what an API key is allowed to do?+
Not through the key itself. The Postman API has no per-endpoint scopes, so a key inherits the full role and team access of the user who created it. The only way to limit an agent is to control whose Postman account the key belongs to, for example creating the key under a user who only has read access to the workspaces the agent should touch. This is the gap Bollard AI closes from the outside, allowing or denying each call per agent.
What are the rate limits?+
An API key is capped at 300 requests per minute overall. A few endpoints are tighter: monitors and the list-collections, list-workspaces, and get-workspace endpoints allow 10 calls per 10 seconds, and workspace updates allow 20 per minute per user. Mock servers allow 120 requests per minute. A monthly allowance also applies per plan. Over the limit, the API returns HTTP 429, and the X-RateLimit headers, including Retry-After, say when to resume.
How do I receive events from Postman?+
Postman does not push events to a subscriber. The closest equivalent runs the other way: the POST /webhooks endpoint creates a webhook whose URL, when called, triggers a collection run through a monitor. The app calls Postman to set this up, and then anything that calls the returned webhook URL kicks off the run; Postman itself does not notify the app when things change inside a workspace.
Does Postman have an official MCP server for AI agents?+
Yes. Postman publishes a first-party Model Context Protocol server that exposes the Postman API as tools an AI agent can call. The remote server answers at mcp.postman.com, with a default minimal tool set, a full set of more than 100 tools, and a code-focused set, plus an EU host at mcp.eu.postman.com. The US remote server supports OAuth and does not require an API key; the EU server uses a Postman API key.
Is the Postman API versioned?+
There is no version number in the API address. Postman ships the API as a public collection that it revises over time, and that collection carries a version label, currently v1.39, which marks the current cut of endpoints. New endpoints are added in later collection versions without breaking earlier calls, so an integration generally does not have to pin or migrate a version.
Related

More developer API guides for agents

What is Bollard AI?

Control what every AI agent can do in Postman.

Bollard AI sits between a team's AI agents and Postman. Grant each agent exactly the access it needs, read or write, area by area, and every call is checked and logged.

  • Set read, write, or full access per agent, never a shared Postman API key.
  • Denied by default, so an agent reaches only what has been explicitly allowed.
  • Every call recorded in plain English: who, what, where, and the decision.
Postman
API Ops Agent
Read collections ResourceOffReadFull use
Run monitors ActionOffReadFull use
Delete workspaces ActionOffReadFull use
Per-agent access, set in Bollard AI, not in Postman