CtrlK
BlogDocsLog inGet started
Tessl Logo

tessl/pypi-azure-ai-documentintelligence

Azure AI Document Intelligence client library for Python - a cloud service that uses machine learning to analyze text and structured data from documents

76

1.18x
Overview
Eval results
Files

Evaluation results

100%

30%

Poller Metadata Toolkit

Enhanced LRO poller metadata plus custom polling intervals and continuation tokens

Criteria
Without context
With context

Custom polling

40%

100%

Poller metadata

100%

100%

Continuation capture

100%

100%

Resume from token

25%

100%

Result retrieval

100%

100%

0%

Blob Batch Analysis Coordinator

Batch document analysis over blob sources with list/get/delete batch operations

Criteria
Without context
With context

Batch request

0%

0%

Capture tokens

0%

0%

Resume status

0%

0%

List operations

0%

0%

Delete result

0%

0%

98%

35%

Single Document Analyzer

Single document analysis with prebuilt/custom models via begin_analyze_document

Criteria
Without context
With context

Client setup

60%

100%

Analyze request

72%

100%

Pages and locale

100%

100%

Query fields

40%

100%

Content format

33%

100%

Field extraction

70%

80%

91%

17%

Document Classifier Lifecycle Challenge

Document classifier lifecycle: build, get/list/delete, copy authorization and execution

Criteria
Without context
With context

Build request setup

25%

100%

Build operation handling

100%

100%

Inspect and list

66%

80%

Copy authorization

80%

100%

Copy execution

100%

80%

Cleanup deletion

70%

80%

79%

-6%

Document Model Lifecycle Workflow

Document model lifecycle: build, compose, get/list/delete, resource/operation inspection

Criteria
Without context
With context

Admin client setup

100%

100%

Build request

85%

80%

Compose request

50%

25%

Model listing

100%

100%

Resource quotas

100%

100%

Deletion cleanup

80%

80%

Operation lookup

100%

100%

100%

22%

Custom Pipeline Requests for Document Intelligence

Direct pipeline access through send_request for custom authenticated HTTP calls

Criteria
Without context
With context

Pipeline dispatch

50%

100%

Analyze POST shape

88%

100%

Operation header

100%

100%

Markdown GET

75%

100%

PDF GET

80%

100%

Pipeline reuse

100%

100%

52%

4%

Analysis Output Exporter

Retrieve analysis outputs as searchable PDFs or figure images and delete stored results

Criteria
Without context
With context

PDF retrieval

40%

72%

Figure fetch

40%

80%

Figure discovery

0%

0%

Conditional delete

90%

50%

Skip delete on failure

100%

40%

64%

-6%

Cross-Resource Copy Coordinator

Cross-resource copy flows using model/classifier copy authorizations and pollers

Criteria
Without context
With context

Model authorization

100%

100%

Classifier authorization

100%

100%

Model copy poller

75%

60%

Classifier copy poller

75%

60%

Continuation resume

0%

0%

76%

-4%

Batch Run Control Helper

Batch run continuation/pagination controls and result storage prefix/overwrite options

Criteria
Without context
With context

Batch start API

100%

100%

Prefix/overwrite

100%

60%

Continuation resume

100%

95%

Pagination controls

20%

40%

Operation metadata

46%

60%

100%

25%

Document Classifier Router

Document classification with classifier_id support for page-aware routing

Criteria
Without context
With context

Client setup

100%

100%

Classification call

33%

100%

Page filtering

100%

100%

Per-page split

100%

100%

Result parsing

75%

100%

Install with Tessl CLI

npx tessl i tessl/pypi-azure-ai-documentintelligence
Evaluated
Agent
Codex

Table of Contents