Azure Document Intelligence pricing

Question

Azure Document Intelligence pricing

Chithra P 40

Hi,

Could you provide the pricing structure used for Azure Document Intelligence

0 comments

2 answers

Your answer

Answer 1

Welcome to Microsoft Q&A

Hello @Chithra P I hope you are doing well, i was going to responde you in the other thread but dont worry , this info might help you.

this is the official guide @Chithra P take a look https://azure.microsoft.com/en-gb/pricing/details/document-intelligence/

Azure Document Intelligence operates on a consumption-based (pay-as-you-go) pricing model. Your total cost is determined by two main factors: the specific AI model you choose to invoke and the total number of pages processed per month.

Pricing Structure Breakdown

Free Tier (F0): Excellent for proof-of-concept testing. You receive an allowance of 500 pages per month for free, but it is strictly limited to processing only the first 2 pages of any given document and a maximum file size of 4 MB.

Read Model (OCR only): The most cost-effective tier, designed for simple text extraction without complex structural understanding. It costs roughly $1.50 per 1,000 pages.

Layout & Prebuilt Models: Used for extracting structure (tables, checkboxes) or leveraging domain-specific models trained by Microsoft (Invoices, Receipts, IDs, W-2s, etc.). This tier costs approximately $10.00 per 1,000 pages.

Custom Models: Used when you need to train the AI on your own highly specialized document templates. Custom Extraction is generally billed at $30.00 per 1,000 pages.

(Note: Training a custom model is completely free; you are only charged for inference when you analyze documents against that model. Additionally, using advanced add-ons like barcode extraction or high-resolution scanning will increase the base unit price of your chosen model.)

Charges are calculated per 1,000 pages processed or per transaction, depending on the model:

Model type	Main function	Approx. price (per 1,000 pages)	Notes
Detects text and table structure	~US $15–20	Best for generic documents
--------	--------	--------
Layout	Detects text and table structure	~US $15–20	Best for generic documents
Prebuilt models (Receipts, Business Cards, IDs)	Extracts data from common document types	~US $100–110	Specialized OCR
Custom Classification	Classifies documents by type	~US $30–35	Requires training
Custom Extraction	Extracts specific fields/values	~US $300–310	For invoices, contracts, forms
Generative Extraction	Advanced AI-driven extraction	~US $305–310	Region availability varies
Model training	Cost per training hour	~US $30–35/hour	Applies only to custom models

Free tier: up to 500 pages/month at no cost. Volume discounts: available for high usage with commitment tiers.

Region: prices vary by currency and subscription type (Commercial, Government, CSP).
Exchange rates: Azure uses London FX rates two days before month-end.
Contract type: Pay‑As‑You‑Go, Enterprise Agreement, or CSP may have different rates.
Azure AI Document Intelligence pricing — Microsoft Learn
Azure Pricing Calculator — estimate costs for your region and workload
Custom Field Extraction overview
https://www.youtube.com/watch?v=LzwblkdxdRw

If my answer helped you resolve your issue, please consider marking it as the correct answer. This helps others in the community find solutions more easily. Thanks!

Answer 2

SRILAKSHMI C 19,550 Microsoft External Staff Moderator

Hello @Chithra P

Thank you for your question regarding Azure AI Document Intelligence pricing.

Azure AI Document Intelligence is a consumption-based service, and billing is primarily calculated based on the number of pages processed. The exact cost depends on the feature or model used (such as prebuilt models, custom extraction models, classification models, or other advanced capabilities), as each feature may have a different pricing rate.

How Azure AI Document Intelligence Pricing Works

Azure AI Document Intelligence offers two pricing options:

Pay-As-You-Go (PAYG) – You are billed based on the actual number of pages processed.
Commitment Tiers – Designed for higher-volume workloads, providing discounted pricing in exchange for committing to a monthly usage volume.

Estimating Your Usage and Costs

To determine your monthly consumption:

Navigate to your Azure AI Document Intelligence resource in the Azure portal.
Select Monitoring > Metrics.
Choose the Processed Pages metric.
Under Values, select Apply Splitting and choose FeatureName.
Review the number of pages processed for each feature.
Multiply the page count by the corresponding unit price listed on the official pricing page.

This provides the most accurate estimate of your monthly costs.

If you are using the Free (F0) tier:

Up to 500 pages per month can be processed at no charge.
Once the free limit is reached, additional processing requires upgrading to a paid tier such as Standard (S0).

Commitment Tier Considerations

If your workload involves large document volumes (for example, tens of thousands of pages per month), Commitment Tiers may offer better cost efficiency than Pay-As-You-Go pricing.

If you notice higher-than-expected costs:

Review the Processed Pages metric to identify usage trends.
Check whether additional features or models were used.
Use Azure Cost Management + Billing to analyze charges by resource and service.
Review Azure Monitor metrics for any unusual spikes in document processing activity.

Please refer to:

Azure AI Document Intelligence pricing explainer source link (pricing page referenced in doc): https://azure.microsoft.com/pricing/details/ai-document-intelligence/

Azure AI Document Intelligence service limits (referenced in doc): https://learn.microsoft.com/azure/ai-services/document-intelligence/service-limits

Quickstart: Purchase commitment tier pricing: https://learn.microsoft.com/azure/cognitive-services/commitment-tier

Pricing for Azure Content Understanding in Foundry Tools (example context referenced): https://learn.microsoft.com/azure/ai-services/content-understanding/pricing-explainer?wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#pricing-example-invoice-field-extraction

“What is Azure Document Intelligence in Foundry Tools?” (context for analysis features/premium features):

https://learn.microsoft.com/azure/ai-services/document-intelligence/overview?view=doc-intel-4.0.0&wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#analysis-features

I Hope this helps. Do let me know if you have any further queries.

If this answers your query, please do click Accept Answer and Yes for was this answer helpful.

Thank you!

SRILAKSHMI C 19,550 Reputation points Microsoft External Staff Moderator

2026-06-25T11:58:03.62+00:00

Hi @Chithra P,

Following up to see if the above answer was helpful. If this answers your query, please do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

Thank you!
Chithra P 40 Reputation points

2026-06-25T14:56:33.02+00:00

While testing the pdf consist of checkbox but the checked check boxes are not correctly extracting. All the checkbox values are return as response
SRILAKSHMI C 19,550 Reputation points Microsoft External Staff Moderator

2026-06-28T07:54:10.85+00:00
Hi @Chithra P,

Thank you for your follow-up.

From your description, I understand that while testing a PDF containing checkboxes, the checked and unchecked checkbox values are not being extracted correctly. Instead, the service appears to be returning the same value for all checkboxes, regardless of whether they are selected.

To help us better understand the behavior, could you please provide the following information?

Which Azure AI Document Intelligence model are you using? (For example, Read, Layout, Prebuilt, or a Custom Extraction model.)

Which API version are you using?

A sample PDF that reproduces the issue (if it can be shared), or a screenshot of the checkbox section.

The JSON response returned by the Document Intelligence API.

The expected output versus the actual output.

Whether the PDF is:

A digitally generated PDF, or

A scanned document.

Checkbox extraction accuracy can vary depending on factors such as:

The quality and resolution of the document.

Whether the checkboxes are clearly marked and distinguishable.

The document layout and formatting.

The model being used for analysis.

If you're using a Custom Extraction model, please also let us know whether your training dataset includes examples containing both checked and unchecked checkboxes, as representative training data can improve extraction accuracy.

Once we have the requested information and a sample document, we'll be able to determine whether the behavior is expected, related to the document format, or requires further investigation by the engineering team.

For reference, you may review the Azure AI Document Intelligence documentation:

Azure AI Document Intelligence overview: https://learn.microsoft.com/azure/ai-services/document-intelligence/overview

Model overview and supported capabilities: https://learn.microsoft.com/azure/ai-services/document-intelligence/model-overview

Thank you!
SRILAKSHMI C 19,550 Reputation points Microsoft External Staff Moderator

2026-06-30T06:08:37.1633333+00:00

Hi @Chithra P,

Following up to see if the above answer was helpful. If this answers your query, please do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

Thank you!
SRILAKSHMI C 19,550 Reputation points Microsoft External Staff Moderator

2026-07-02T01:00:01.5533333+00:00

Hi @Chithra P,

Just checking in to see if you have got a chance to see my response to your question in resolving the issue.

If you are still facing any further issues, please don't hesitate to reach out to us. We are happy to assist you.

Looking forward to your response and appreciate your time on this.

If you feel that your quires have been resolved, please accept the answer by clicking the "Upvote" and "Accept Answer" on the post.

Thank you!