Azure OpenAI and Azure AI Foundry cost tracking, by model and deployment type

What CloudQuell reads

Foundry line items from the Azure cost data your integration already reads. Each meter's name and subcategory say which model ran, who built it, what kind of usage it billed, and how it was deployed.

Model

Parsed from the meter name and subcategory, so "5.4 mini Opt Gl 1M Tokens" reads as GPT-5.4 Mini and "R1 Inp glbl Tokens" as DeepSeek R1, with spend by model, a model cost comparison table, and spend over time stacked by model.

Model developer

Who built each model (OpenAI, DeepSeek, xAI, Meta, Mistral AI, Microsoft and others), read from the meter subcategory and shown as its own spend breakdown.

Token type

Input (uncached), output, cache read, and cache write, counted only on meters that bill tokens. Provisioned throughput (PTU), fine-tuning training and hosting, and Marketplace Claude are kept as cost without token counts, so they never skew the per-token math.

Deployment type

Global, data zone, or regional, read from the same meter, with each deployment type’s spend and share of the total.

Subscription, region, day

Foundry spend by subscription, by Azure region, and by day or month, with a Subscription filter to narrow the whole view.

History

The view covers whatever Azure history the integration holds: up to 13 months on quick connect, or as far back as your one-time FOCUS exports go.

What the data looks like in AI Spend

One Foundry meter, decoded. The meter name carries the model, the token type, and the deployment type; the subcategory names the model family and its developer.

How it connects: via your Azure connection

There is no Azure OpenAI connector. If CloudQuell can read your Azure bill, it can read your Foundry spend.

Connector

Your Azure integration: the read-only Entra ID service principal, on the FOCUS export or quick-connect path. Nothing extra to install, grant, or key in.

When it appears

An Azure AI Foundry view turns on in AI Spend by itself once your Azure bill carries Foundry model spend. Azure reports usage with an 8–24 hour lag, and the integration syncs daily at 07:30 UTC.

Counted once

Foundry spend stays inside your Azure cloud spend. AI Spend is a read-only lens over those same rows, and its All AI total is never added to cloud totals, so the Overview, budgets, and allocation still reconcile to your Azure invoice.

Cost basis

Foundry figures follow the cost settings on your Azure integration, the same as every other Azure line.

Plans

AI Spend, Azure AI Foundry included, is on every plan. Free covers up to $10K of monthly cloud spend. Alerts and budgets are on paid plans.

What it unlocks

Azure AI Foundry gets the same treatment as the rest of your AI bill, without being counted twice.

• All your AI spend, side by side

  Switch AI Spend between All AI, Anthropic, OpenAI, Bedrock, Azure AI Foundry, and Gemini / Vertex, for the sources you have. Under All AI, a GPT model on Azure and the same model on the OpenAI API share one model line.

  OpenAI integration →

• Effective $ per 1M tokens

  A blended rate over token-metered spend only, with a note on how much of your Foundry bill it covers. PTU, fine-tuning, hosting, media, and Marketplace Claude stay out of the rate. Cache-read share and a projected month-end sit beside it.

• Alerts, budgets, and allocation

  Foundry spend is Azure spend, so per-service anomaly detection, budgets scoped to a subscription, service, or tag, and alerts to Slack, Microsoft Teams, or email cover it on paid plans. Route it to cost centers with the same tag and service rules as the rest of Azure.

  Azure integration →

• From an AI agent

  On paid plans, the MCP tool get_ai_spend returns the same rows with provider "azure_openai", grouped by model, model developer, token type, deployment type, subscription, region, service tier, or day. Cloud-billed sources are off by default there, so an agent can’t add them to a cloud total by accident.

  FinOps MCP server →

Common questions

Do I need a separate integration to track Azure OpenAI costs?

No. Azure OpenAI is billed by Microsoft, not OpenAI, so its charges are part of your Azure bill and arrive through your existing Azure connection: the same read-only service principal, on the FOCUS export or quick-connect path. Once that spend appears, an Azure AI Foundry view turns on in AI Spend by itself. There is nothing extra to install or grant, and the OpenAI integration is not involved.

Can I see Azure OpenAI cost per deployment?

By deployment type, yes; by named deployment, no. CloudQuell splits Foundry spend into global, data zone, and regional deployments, and by model, subscription, and region, all read from the billing meter. It does not break spend out by the deployment names you create. For one named deployment, Azure Cost Management can split a Foundry resource’s cost across its model deployments in the cost-by-resource view.

Can Azure Cost Management track Azure AI Foundry costs by itself?

For Azure alone, largely yes. Microsoft documents grouping Foundry costs by meter and by resource, filtering by deployment tags, a preview project tag for chargeback on models sold by Azure (not yet for Azure Marketplace models), and budgets with alerts. Azure OpenAI usage sits under the broader Cognitive Services service classification, so Microsoft points you to a service tier filter instead of a service filter. CloudQuell adds the cross-provider view: meters decoded into model, token type, and deployment type, beside your direct OpenAI, Anthropic, Bedrock, and Gemini / Vertex spend, with alerts to Slack, Microsoft Teams, or email.

Is Azure AI Foundry spend double-counted with my Azure total?

No. Foundry spend is counted once, inside your Azure cloud spend. AI Spend is a read-only lens over those same rows: its All AI figure combines direct and cloud-billed AI spend for analysis and is never added to cloud totals, budgets, or allocation, so every number still reconciles to your Azure invoice.

Which Azure AI costs show up in AI Spend?

Model spend: Azure OpenAI, the other models Microsoft sells through Foundry (DeepSeek, Grok, Llama, Mistral, Phi, Microsoft MAI), and Claude bought through Azure Marketplace. Provisioned throughput (PTU) and Marketplace Claude appear as their own buckets rather than per model; Microsoft bills Marketplace Claude as a single line with no per-model split. Fine-tuning training and hosting count as that model’s spend, without token counts. Azure Speech, Translator, and Vision are not model spend and stay in your Azure cloud spend only.

What does Azure OpenAI cost tracking cost?

AI Spend, Azure AI Foundry included, is on the Free plan for up to $10K of monthly cloud spend, with Azure counting as one of three connected integrations. Paid plans are a flat monthly fee: Starter $99/month up to $50K of monthly spend, Growth $199/month up to $200K, with a 14-day trial and no credit card. Never a percentage of your bill.