AI Token Spend Management

Overview

AI Token Spend Management gives your finance team consolidated visibility into token usage, costs by model, and spend trends across AI providers like Anthropic and OpenAI. Instead of reconciling invoices from multiple dashboards, you can track it all in one place on Ramp.

• See a breakdown of costs by provider, model, team, and users
• Track token usage and identify spend trends across your organization
• Spot cost spikes and understand what's driving them
• Compare cost-per-token efficiency across providers and models

Note: Currently, only Admins, Auditors, Finance Admins, and IT Admins can access this product.

Getting started

Who can access AI Token Spend Management

AI Token Spend Management is available to users with the following roles: Admin, Auditor, Finance Admin, and IT Admin. We will be expanding access to other roles down the line.

Where to find it

In the Ramp app: Early Access → enable AI Token Spend Management → open the AI Spend page.

Connect your AI providers

Connect your AI provider so Ramp can load your cost and usage data. For Anthropic and OpenAI, this requires an Admin API key (typically provided by your IT Admin). For GCP, Ramp connects through a BigQuery billing export table ID instead.

To connect Anthropic Platform

1. Go to platform.claude.com/settings/admin-keys and sign in as an admin
2. Navigate to Settings > Organization > Admin API Keys
3. Create a new key (it will start with sk-ant-admin... )
4. Name it something descriptive, like platform_analytics_ramp
5. Copy and paste it in AI Spend page

To connect Claude Enterprise

1. Go to claude.ai/analytics/api-keys and sign in as a Primary Owner of your Enterprise organization
2. Navigate to Analytics > API Keys
3. Find the Access toggle under Analytics API and toggle it on
4. Click Create key and add read:analytics , read:spend_limits , and write:spend_limits scopes
5. Copy and paste it in AI Spend page

To connect OpenAI Platform

1. Go to platform.openai.com/settings/organization/admin-keys and sign in as an Organization Owner
2. Navigate to Settings > Organization > Admin API Keys
3. Create a new key (it will start with sk-admin... )
4. Name it something descriptive, like platform_analytics_ramp
5. Copy and paste it in AI Spend page

To connect ChatGPT Enterprise

1. Go to platform.openai.com/settings/organization/api-keys
2. Click Create new secret key . Name it something descriptive, like enterprise_analytics_ramp , and scope it to read_only . Copy the secret key value when OpenAI displays it and store it in your organization's secure credential manager; OpenAI will not show it again.
3. Open the ChatGPT Admin console at chatgpt.com/admin and copy the Workspace ID from the Workspace Settings page
4. Email [email protected] . Provide your Workspace ID and the name of the key you created in Step 2. Ask OpenAI to add the codex.enterprise.analytics.read scope to that key. Do not send the secret key value by email.
   Note: This step is a requirement from OpenAI.
5. After OpenAI confirms the scope has been added, return to the AI Spend page and paste the saved key

To connect GCP - Gemini

You need the following Google Cloud permissions before starting:

• Billing Account Costs Manager or Billing Account Administrator on the target Cloud Billing account
• BigQuery User on the project that will contain the dataset
• Project Creator (only if you need to create a new project)

Step 1: Select or create a Google Cloud project

Before enabling Cloud Billing data exports, choose or create a Google Cloud project where the exported billing data will be stored. For ease of management, we suggest creating a new project specifically for billing if one doesn't already exist. Make sure billing is enabled on this project and that it is linked to the Cloud Billing account that pays for your Gemini or Vertex AI projects — otherwise the export setup will not offer this project as a destination.

Step 2: Create a BigQuery dataset

1. Open BigQuery in the Google Cloud console.
2. Select the project you created or chose in Step 1 from the project selector at the top of the BigQuery console.
3. Click Create dataset.
4. Enter a dataset ID, such as all_billing_data.
5. Click Create dataset. The dataset should now appear under your selected project.

Step 3: Enable Cloud Billing export

1. Click the navigation menu icon to the left of the Google Cloud logo, then select Billing.
2. If you have more than one Cloud Billing account, select the billing account that pays for your Gemini or Vertex AI projects. The export covers only the projects paid by the selected account.
3. Click Billing export.
4. Click Enable detailed export.
5. Complete the setup modal with the project you want to export billing data to and the dataset you created in Step 2.

Step 4: Copy the billing table ID into Ramp

1. The billing export should automatically create a BigQuery table under your selected project and dataset. This may take about 24 hours.
2. After the table is created, navigate to it, select the three-dot menu next to it, and click Copy ID.
3. Return to the AI Spend page in Ramp, paste the billing table ID, and save the connection.

If you have multiple keys per provider

• Settings → Manage → Add API key
• Once connected, the AI Spend page will load with the additional data
  

Data and sync details

Security and data handling

Ramp takes the security of your provider credentials and usage data very seriously.

Credential storage

For Anthropic and OpenAI, Admin API keys are stored in an encrypted format consistent with Ramp's existing integration security standards. Decrypted keys are not stored by Ramp; we decrypt your key when the system makes a call to the provider's admin APIs. For GCP, Ramp stores only the BigQuery billing table ID you provide — Ramp does not store any GCP credentials.

Access control

Only users with Admin, Auditor, Finance Admin, or IT Admin roles can view AI Spend data. Non-admin visibility is not currently supported.

Data minimization

AI Token Spend Management is designed to ingest the fields needed to power spend reporting and analytics and does not require you to provide prompts or message content.

How the product works

For Anthropic and OpenAI, AI Token Spend Management uses the Admin API keys you provide to call your provider's Admin APIs and import usage and billing-related data (e.g., token usage, cost, provider/model identifiers, and any provider-provided dimensions available via the Admin APIs). The Admin API key will have additional access due to platform limitations; however, we only use it to read usage and cost data needed for AI Token Spend Management and will provide notice in advance of any changes so that you can adjust your settings if you wish. For GCP, Ramp reads cost and usage data directly from the BigQuery billing export table you configure — no API key is involved.

Retention

Ramp retains imported spend/usage data to provide historical reporting and trend analysis. If you disconnect a provider, Ramp will cease imports from that provider. For Anthropic and OpenAI, the Admin API keys will be immediately hard deleted from our storage. For GCP, the stored billing table ID will be deleted. Any retained usage data will follow Ramp's standard retention and deletion practices.

Controlling your access

For Anthropic and OpenAI, you can revoke the Admin API key you provided to Ramp at any time to block access to your data. Key revocation can be done from both the OpenAI Console and the Claude Console. Both OpenAI and Anthropic maintain audit logs that show certain activity performed by your Admin API keys. In most cases, you can verify Ramp's activity directly from these audit logs. For GCP, you can revoke Ramp's access by removing the billing table ID from Ramp or revoking Ramp's read permissions on the dataset in the Google Cloud console.

Frequently asked questions

What data does Ramp import from my AI providers?

Ramp imports usage and cost data exposed through the provider Admin APIs you connect. For example: token usage, spend/cost, provider/model identifiers, and time-based usage metrics. This data is used to power spend reporting and analytics in AI Token Spend Management.

Does Ramp ingest prompts, message content, or model outputs?

AI Token Spend Management is designed for spend and usage reporting and does not require prompts or message content. Ramp uses provider Admin APIs to import usage and cost information rather than application content.

Is any of this personal data?

Typically, the imported data is business account usage and cost data. Depending on your provider configuration, the usage export may include identifiers like a user name, email address, or internal user/team label. If present, Ramp uses those identifiers only to provide spend breakdowns and access-controlled reporting.

How does Ramp use this data?

Ramp uses imported usage/cost data to provide reporting, trend analysis, and cost allocation insights inside AI Token Spend Management. Aggregated or anonymized information may be included in Ramp reports or benchmarks, unless this is precluded by your company's agreement with Ramp. Other Ramp customers do not see your company's provider-level, model-level, team-level, user-level, or API key-level data.

Can Ramp employees see my usage data?

Access is restricted and role-based. Only authorized personnel may access customer data for limited purposes such as support, troubleshooting, and maintaining the service, consistent with Ramp's access controls and logging practices.

Can I disconnect a provider or revoke access?

Yes. For Anthropic and OpenAI, revoke the Admin API key in your provider console or disconnect the provider from Ramp. For GCP, remove the billing table ID from Ramp or revoke read permissions on the BigQuery dataset. After disconnecting any provider, Ramp will stop future imports.

How long does Ramp retain this data?

Ramp retains imported data to provide historical reporting and trend analysis. Data retention and deletion follow Ramp's standard practices and contractual obligations.

Does Ramp use this data to train AI models?

AI Token Spend Management uses your provider data to display reporting and analytics. Ramp does not use your company's provider credentials to perform any actions in your provider account beyond retrieving usage and cost data in connection with AI Token Spend Management.

Can other Ramp customers see my company's AI usage data?

No. Other customers do not see your company's provider-level, model-level, team-level, user-level, or API key-level data.

How does Ramp aggregate and de-identify AI spend data for benchmarks?

Ramp may use aggregated and anonymized customer data to report industry-level AI spending trends and benchmarks. Benchmark reporting is not intended to identify your business, individual users, API keys, or provider account details.

Can I give non-admin users access to AI Token Spend Management?

Not currently. Only users with Admin, Auditor, Finance Admin, or IT Admin roles can view the AI Token Spend Management tab.

Why is my data taking a long time to load?

Anthropic rate limits are currently 1 request per minute, so larger accounts may take longer to import. We recommend waiting overnight for the initial data load to complete in the background.

How are costs calculated?

Ramp pulls both cost and usage data from your API provider and cross-references them. If you have custom pricing or credits with your provider, reach out to us so we can reflect those in the calculation.

How often does data sync?

Data syncs daily and should be up to date with T-1 day.

What permissions are needed?

For Anthropic and OpenAI, the Admin API key should be limited to read-only permissions for the provider endpoints required to retrieve usage and billing data. For GCP, you need Billing Account Costs Manager or Billing Account Administrator on the billing account, plus BigQuery User on the dataset project. Ramp uses these credentials only to retrieve the usage and cost data needed for AI Token Spend Management.