> For clean Markdown of any page, append .md to the page URL. > For a complete documentation index, see https://docs.fabricgateway.ai/fabric/administration/llms.txt. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.fabricgateway.ai/_mcp/server. # Manage access > Manage who has access to your Fabric AI Gateway organization and which credentials the gateway can use on your behalf. > **Info** > > This topic describes the product direction and implementation, but much of the platform is still evolving. Some details, capabilities, and decisions may change as development continues. **Admin** is where you manage who has access to your Fabric AI Gateway organization and which credentials the gateway can use on your behalf. Use this section to add or remove users, change roles, invite team members, store credentials, and review activity. What you see here depends on your role and on how your data plane is hosted. This topic explains both. ## Administration guides | Section | What it's for | Guide | | --------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------- | | **Members & Invites** (**Members** tab) | Identities provisioned into your organization (for example, through an identity provider such as Okta), along with their role and status. | [Manage users and invitations](/fabric/administration/users) | | **Members & Invites** (**Invites** tab) | Invitations that you have sent but people haven't yet accepted. | [Manage users and invitations](/fabric/administration/users) | | **Credentials** (main sidebar) | The credentials that the gateway uses and issues for your connected services. | [Manage credentials and vaults](/fabric/administration/credentials) | | **API keys** | Keys for managing your Fabric AI Gateway organization programmatically. | [Manage API keys](/fabric/administration/api-keys) | | **Namespaces** | Separate teams' connections, services, analytics, traces, and users within one organization. | [Manage namespaces](/fabric/administration/namespaces) | | **Audit logs** | Security and activity events, and LLM token metering. | [Review audit logs](#review-audit-logs) | | **Usage & Billing** | Your plan and the requests that each connection uses in the current period. | [Review usage and billing](#review-usage-and-billing) | Enterprise tenants also include the following items: | Section | Where | What it's for | | ------------------ | ----- | --------------------------------------------------------------------------------------------------------------------- | | **Single sign-on** | Admin | Configuration for how users are provisioned into the organization. | | **Vaults** | Admin | The vaults where you store your own secrets. See [Manage credentials and vaults](/fabric/administration/credentials). | | **Gateway health** | Admin | Health status for the data plane. | ## Understand cloud and enterprise tenants Fabric AI Gateway has two deployment models, and the console differs between them: * **Cloud** — Postman hosts both the control plane and the data plane, so your tenant lives in Postman's cloud and uses a shared data plane. * **Enterprise** — This is a hybrid deployment. Postman hosts the control plane, and the data plane runs in your own cloud. Your tenant is dedicated. The following table shows the differences you see in the console: | Item | Cloud tenant | Enterprise tenant | | ------------------ | ----------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | | **Vaults** | Not shown. Fabric uses a Postman-managed vault in the background. | Shown. You can bring your own vault, in your own cloud, so your secrets stay in your environment. | | **Single sign-on** | Not shown. | Shown. | | **Gateway health** | Not shown. | Shown. | If you don't see an item described in these docs, check which kind of tenant you're in. For more about each model, see [Deployment options](/get-started/architecture#deployment-options). ## Understand roles Each user has one of two roles: | Role | What the user sees | | --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Super admin** | Everything. | | **Member** | The **Gateway** items (read access to **LLMs**, **MCP servers**, **Services**, and **Plugins**), **Members**, and **Credentials**. Members don't see **Admin**, which includes **Analytics** and **Traces**. | To change a user's role, see [Manage users and invitations](/fabric/administration/users). ## Switch organizations If you belong to more than one organization, you can switch between them without signing out. Each organization is an isolated organization. To switch organizations, do the following: 1. At the bottom of the sidebar, select your organization name. A menu opens that includes **Account settings**, **Theme**, **Help**, **Switch organization**, and **Log out**. 2. Click **Switch organization**, and then select the organization you want to open. A check mark shows the organization you're in. ## Review audit logs **Audit logs**, under **Admin**, is where you review what's happening in your organization. It brings together two views: security and activity events, and token metering for LLM traffic. Use the first to see who did what and whether it worked, and the second to see how many tokens your LLM requests use and what they cost. To review traffic and cost trends across all of your routes, see [Review analytics](/fabric/observability/analytics). ## Review usage and billing **Usage & Billing**, under **Admin**, shows your plan and how much each connection uses. It appears when billing is turned on for your organization, for example on a Postman-hosted data plane. The page includes the following: | Section | What it shows | | --------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Plan | Your current plan and its price. For example, the **Free plan** shows `$0` and explains that API, MCP, and LLM requests are unlimited and included at no charge. | | **Connections** | One row for each connection, with its **Type** (**LLM** or **MCP**), the **Requests this period**, and its **Allowance**. The allowance is **Unlimited** on the Free plan. | If the Free plan doesn't meet your needs, select **Talk to us** to contact Postman about other plans. For plan details, see [Deployment options](/get-started/architecture#deployment-options). ## Sign in to Fabric AI Gateway To sign in, do the following: 1. Go to your organization's sign-in page. 2. Enter your email address. 3. Click **Continue with email**. If you don't have an account yet, click **Sign up** on the sign-in page instead. ## Next steps * To connect your first provider before you invite your team, see the [Quickstart](/get-started/quickstart). * To control what traffic does after people have access, see [Control traffic with plugins](/fabric/plugins). > Manage who has access to your Fabric AI Gateway organization and which credentials the gateway can use on your behalf. ## Docs - [Manage users and invitations](https://docs.fabricgateway.ai/fabric/administration/users.md): Review users, invite people, check pending invitations, and change a user's role. - [Manage credentials and vaults](https://docs.fabricgateway.ai/fabric/administration/credentials.md): Manage the credentials that the gateway uses, and understand where secrets are stored on cloud and enterprise tenants. - [Manage API keys](https://docs.fabricgateway.ai/fabric/administration/api-keys.md): Create admin API keys to manage your Fabric AI Gateway organization programmatically. - [Manage namespaces](https://docs.fabricgateway.ai/fabric/administration/namespaces.md): Separate a team's connections, services, analytics, traces, and users into namespaces within one organization.