Document AI for tabular data

has a clear lead for extracting tabular data from PDF financial reports.

It depends

is recommended for table parsing in Microsoft environments.

Since Jul 6Updated Nov 20, 2026Based on 58 AI answersReviewed by

How we measure

Brands AI recommends

Mentioned in, on ChatGPT and Google AI Mode

Key takeaways for Document AI for tabular data

  • Top contenders

    Google Cloud Document AI and Azure AI Document Intelligence tie at #1 for Document AI for tabular data.

  • Across AI Document Parsing & Extraction APIs

    Within AI Document Parsing & Extraction APIs, LlamaParse leads Document table extraction, showing AI shifts picks by buyer use case. See Document table extraction

Questions and answers

  • ChatGPTNov 9, 2026

    1. 1Affinda
    2. 2Docsumo
    3. 3Reducto
    4. 4DocuClipper
    5. 5Infrrd
    6. 6Google Cloud Document AI
    7. 7Azure AI Document Intelligence
    8. 8Amazon Textract
    9. 9SirionOne

    If you're struggling to extract tables from PDF financial reports, I would focus on specialized Document AI and Intelligent Document Processing (IDP) providers rather than conventional OCR software.

    Several vendors specialize in extracting structured tabular data, preserving rows, columns, headers, financial values, and relationships between cells.

    1. Specialized Document AI providers

    These are the vendors I would shortlist.

    • Find a Member — AFIA

    1. Affinda

    Specializes in financial document processing, including annual reports and financial statements.

    • Financial data extraction
    • Structured document parsing
    • Automated processing workflows

    Website:


    • Docsumo fund | Nepali Times

    2. Docsumo

    Particularly relevant for financial documents and complex transaction tables.

    • Extracts tables from scanned and digital PDFs
    • Financial document processing
    • Excel, CSV and JSON exports
    • API integration

    Website:


    • Product Marketing Manager @ Reducto

    3. Reducto

    AI-native document parsing platform for complex PDFs.

    • Table extraction
    • Structured output
    • Document understanding
    • API-based processing at scale

    Website:


    • QuickBooks App Store

    4. DocuClipper

    Focuses on financial data extraction.

    • PDF-to-Excel conversion
    • Financial statements and bank statements
    • Transaction table extraction

    Website:


    • Infrrd - Crunchbase Company Profile & Funding

    5. Infrrd

    Enterprise document processing for complex layouts.

    • Table and field extraction
    • Intelligent OCR
    • Automated document classification
    • Enterprise workflows

    Website:


    • AWS Marketplace: Eigen - Intelligent Document Processing & Data Extraction

    6. Sirion Eigen

    Particularly interesting for extracting entire tables or selected cells.

    • Table classification
    • Cell-level extraction
    • Export to XLSX, CSV and JSON

    Website:

    parse.glwww.sirion.ainanonets.com

    2. Major cloud Document AI platforms

    If you want to build your own automated extraction pipeline, consider these three established providers.

    • Google Cloud Document AI with API | Generative AI Projects | Codersarts AI

    Google Cloud Document AI

    Table extraction, document layout recognition and structured outputs.

    Official website

    • Analyze Forms and Documents with Azure AI Document Intelligence | CloudLearn

    Azure AI Document Intelligence

    Layout analysis, table recognition and custom extraction models.

    Official website

    • Data extraction from documents made easy with Amazon Textract - DEV Community

    Amazon Textract

    Extracts tables, cells, forms and document structure through APIs.

    Official website

    parse.glparseur.comwww.ciopages.com

    3. Comparison: which should you investigate?

    ProviderParticularly suitable for
    AffindaFinancial reports
    DocsumoFinancial document extraction
    ReductoComplex PDF parsing
    DocuClipperFinancial tables to Excel
    InfrrdEnterprise document processing
    Sirion EigenSpecific table and cell extraction
    Google Document AICloud-scale extraction
    Azure Document IntelligenceMicrosoft-based workflows
    Amazon TextractAWS-based pipelines

    4. What capabilities matter for financial reports?

    This is particularly important because financial reports contain structures that conventional OCR frequently misinterprets.

    I would test each provider against these requirements:

    CapabilityWhy it matters
    Multi-page table reconstructionTables often continue across pages
    Merged-cell recognitionFinancial headers have multiple levels
    Negative number recognitionParentheses indicate negative values
    Footnote preservationNotes can materially change interpretation
    Column alignmentPrevents values entering incorrect periods
    Excel/CSV/JSON exportEnables downstream financial analysis
    Source traceabilityVerify extracted numbers against original pages

    www.pdftableocr.comparseur.comwww.sirion.ai

    5. My recommendation

    I would start with Docsumo, Reducto, Affinda and Sirion Eigen.

    They cover different approaches to financial document processing and table extraction. Compare them using the same financial reports rather than relying on advertised OCR accuracy.

    One important question will help me narrow the recommendations:

    What kind of PDF financial reports are you trying to extract?

    1. Type of financial reports

    Annual reports and SEC 10-K filings

    Quarterly financial statements

    Bank statements and transaction reports

    Investment research and analyst reports

    Mixed financial documents


    2. What makes the PDFs difficult?

    Scanned PDFs and image-based tables

    Digitally generated PDFs with complex tables

    Multi-page tables with merged headers

    A mixture of all three


    3. Desired output

    Excel / CSV

    JSON through an API

    Structured database

    Financial analysis workflows

    [Submit requirements]

    I can then narrow this down to the most suitable specialized vendors, including solutions specifically designed for extracting financial statement tables rather than just recognizing PDF text.

    Google AI ModeNov 9, 2026

    1. 1LlamaParse
    2. 2LlamaIndex
    3. 3Docling
    4. 4IBM Research
    5. 5TableFormer
    6. 6UiPath Document Understanding
    7. 7Hyperscience
    8. 8Rossum
    9. 9AWS
    10. 10Landing AI

    Extracting complex, multi-page financial tables (like balance sheets, income statements, or cash flow footnotes) requires layout-aware document AI rather than standard OCR, which often scrambles rows, columns, and negative number formatting.

    Several specialized platforms and open-source tools excel at financial table extraction:

    Top Document AI & Table Extraction Tools

    • LlamaParse (by LlamaIndex): Built specifically for RAG (Retrieval-Augmented Generation) and complex document parsing, it features layout-aware and multimodal extraction designed to preserve complex table structures, multi-line headers, and nested financial data . Available via LlamaIndex Insights.
    • Docling (Open-Source by IBM Research): Utilizes deep-learning models (like TableFormer ) to convert dense financial and scientific PDFs into clean Markdown or structured JSON. Ideal if you need a privacy-sensitive, self-hosted open-source pipeline. Explore via LlamaIndex Insights.
    • LandingAI ADE (Agentic Document Extraction): Uses a two-step parse-and-extract framework (/parse to clean markdown, /extract mapped to a custom JSON schema), making it strong at pulling line items from brokerage statements and balance sheets without rigid templates. Learn more via LandingAI Developers.
    • Enterprise IDP Platforms: If you are looking for enterprise-grade, out-of-the-box automation (SOC2 compliant with built-in human-in-the-loop verification for audit trails), tools like UiPath Document Understanding, Hyperscience, Rossum , and AWS Textract (with Queries/Tables) are widely deployed in corporate finance.

    Are you looking to:

    • Build a custom script or RAG pipeline using a developer API?
    • Set up a no-code enterprise solution with a user interface?
    • Process a massive batch of files locally for data privacy?
  • Google AI ModeOct 21, 2026

    1. 1Google Cloud Document AI
    2. 2Amazon Textract
    3. 3Azure AI Document Intelligence
    4. 4LlamaParse
    5. 5DataSnipper

    The best Document AI tools for extracting complex tables from PDF financial reports include enterprise cloud APIs and developer-focused parsing frameworks . Top options vary by technical setup and workflow needs:

    Enterprise Cloud APIs

    • Google Cloud Document AI : Uses advanced layout and custom foundation models to accurately identify grid structures, key-value pairs, and dense financial tables.
    • Amazon Textract : Excellent at extracting tables and forms from complex digital or scanned financial documents without requiring custom templates.
    • Microsoft Azure AI Document Intelligence : Strong prebuilt layout models that output structured JSON and markdown representations of multi-page financial tables.

    Developer & Open-Source Parsers

    • LlamaParse : Built specifically for retrieval-augmented generation (RAG) pipelines, highly optimized to parse complex financial tables into structured markdown or JSON.
    • Docling : An open-source document conversion tool utilizing TableFormer-based extraction tuned cleanly for structural grids in financial and research reports.
    • DataSnipper : Popular choice for auditors working directly inside Excel to reconcile financial data from PDFs.

    If you can share whether you need a no-code interface , an API for an engineering pipeline , or integration with Excel , I can recommend the ideal specific tool for your project.

    ChatGPTOct 21, 2026

    1. 1Amazon Textract
    2. 2Azure AI Document Intelligence
    3. 3camelot-py
    4. 4ABBYY FineReader PDF
    5. 5ABBYY Vantage
    6. 6Google Cloud Document AI
    7. 7Microsoft Table Transformer

    For extracting tables accurately from PDF financial reports, I’d shortlist Google Document AI, Amazon Textract, Azure AI Document Intelligence, ABBYY, and Docling/Camelot. The best choice depends on whether your reports are digitally generated or scanned, how complex the tables are, and whether you need a no-code tool or an automated pipeline.

    One important distinction: extracting the words and numbers correctly is not the same as reconstructing the table correctly. Financial reports often contain multi-level headers, negative values in parentheses, footnotes, merged cells, and tables that continue across pages. You need a tool that preserves the relationships between rows and columns, not just one that reads the text.

    Best options at a glance

    ToolBest forMain advantageMain trade-off
    Google Document AICloud-scale financial report processingStructured extraction and integration with Google CloudRequires cloud setup and testing
    Amazon TextractHigh-volume automated pipelinesTable extraction API, cell-level structure, AWS integrationMay require post-processing
    Azure AI Document IntelligenceMicrosoft-centric organizationsLayout analysis, OCR, custom extractionCloud integration and usage costs
    ABBYY FineReader PDF / ABBYY VantageDesktop conversion or enterprise workflowsOCR and document-processing capabilitiesEnterprise automation may require a quote
    DoclingOpen-source, local or self-hosted workflowsDocument layout and table structure extractionMay need engineering and validation
    CamelotDigital PDFs with complex financial tablesFine-grained table extraction controlsMore hands-on setup; scanned PDFs need OCR

    The cloud options are strong candidates for scalable extraction; Camelot and Docling are worth testing if you want more control or prefer open-source tooling.

    My top recommendations

    • Evaluate Document Parsers for Complex Documents | Extend

    1. Google Document AI — best cloud option to evaluate

    Best if you process large numbers of reports and want structured output for analytics or data pipelines. It offers document-specific processors, OCR and integration with Google Cloud services. Test its table output on multi-page financial statements rather than assuming general OCR accuracy guarantees accurate tables.

    Best for: recurring extraction at scale.

    Official website and documentation


    • Data extraction from documents made easy with Amazon Textract - DEV Community

    2. Amazon Textract — best for automated table extraction

    Its document-analysis API detects tables and returns structured information about cells and their contents. That makes it a practical candidate for extracting balance sheets, income statements and cash-flow tables into downstream data systems.

    Best for: AWS-based pipelines and batch processing.

    Official website and documentation


    • Analyze Forms and Documents with Azure AI Document Intelligence | CloudLearn

    3. Azure AI Document Intelligence — best for Microsoft environments

    A good choice if your data pipeline already uses Azure, Microsoft Fabric or Power Automate. It provides layout extraction, OCR and custom document-processing capabilities. Check how well it preserves multi-level headers and column alignment in your specific reports.

    Best for: Microsoft-based enterprise workflows.

    Official website


    • ABBYY FineReader PDF 16 | The Document Warehouse

    4. ABBYY FineReader PDF — best desktop-oriented option

    Worth trying if analysts need to convert reports into Excel without building an API integration. ABBYY also offers Vantage for enterprise document processing. Compare the exported spreadsheets carefully for merged cells, numeric formats and footnotes.

    Best for: analyst-led PDF-to-Excel workflows and OCR.

    FineReader PDF ·


    • Alibaba Cloud OCR官网入口、公司介绍与详细资料 - AIBestBox 智模导航

    5. Docling and Camelot — best open-source candidates

    Consider these if you want to run extraction in your own environment and control the processing pipeline. Camelot is particularly useful for digitally generated PDFs, with different parsing strategies for bordered and whitespace-separated tables. Docling is worth testing for broader document-layout processing. A comparative study also found that Microsoft's Table Transformer performed particularly well at table detection in financial documents.

    Best for: Python workflows, customization and local processing.

    Docling ·

    How to choose for financial reports

    Digitally generated annual reports and 10-Ks

    Start with Camelot and Docling, then compare with a cloud service if the tables are unusually complex.


    Scanned reports or image-only PDFs

    Start with Google Document AI, Textract or Azure Document Intelligence. OCR quality becomes a major part of the result.


    Complex tables with merged headers or multiple pages

    Test structure preservation explicitly. Camelot's current documentation describes handling of merged headers and multi-page table stitching; model-based table detection is another candidate.


    Confidential financial data

    Check retention policies, training-data use, regional processing and deployment options. If documents cannot leave your environment, prioritize local processing and verify the full pipeline's dependencies.

    How I would test accuracy before buying

    Don't rely on a vendor's advertised accuracy or a generic benchmark. Run the same 20–50 representative reports through your top two or three candidates and compare:

    • Cell-level accuracy: Are every row, column and numeric value correct?
    • Financial fidelity: Are negatives, parentheses, decimal places, percentages and currency units preserved?
    • Table structure: Are multi-level headers, blank cells, subtotals and footnotes represented correctly?
    • Multi-page continuity: Are continued tables combined without dropping or duplicating rows?
    • Traceability: Can you map extracted figures back to their source page and location?
    • Total cost: Include API charges, implementation, exception handling and human review.

    For financial use, I would also validate extracted totals against the source statements and flag any arithmetic discrepancies for review. A tool can produce a neatly formatted spreadsheet that still contains a materially incorrect figure.

    One question to narrow this down

    Are you looking for a ready-to-use tool where you upload PDFs and export to Excel, or an AI/API solution to extract tables automatically from hundreds or thousands of reports?

    That distinction would change my shortlist considerably.

  • Google AI ModeNov 18, 2026

    1. 1LlamaIndex
    2. 2Docling
    3. 3IBM Research
    4. 4Azure AI Document Intelligence
    5. 5Google Cloud Document AI
    6. 6Nanonets

    Extracting tabular data from complex financial reports (such as 10-Ks, balance sheets, and cash flow statements) requires tools that preserve cell alignment, handle multi-page spanning tables, and understand nested headers rather than flattening them into messy text.

    Top document AI solutions specialize in financial tabular extraction:

    Top Recommendations

    • LlamaParse (by LlamaIndex) : A premier choice for complex financial documents and RAG (Retrieval-Augmented Generation) workflows . It handles intricate layouts, preserves table structures into Markdown or JSON, and offers schema-guided extraction. Learn more on the LlamaIndex Overview Page.
    • Docling (by IBM Research / Open Source) : An open-source, local-first powerhouse designed to convert complex PDFs—including dense financial tables—into structured formats (JSON, Markdown, or HTML) with high fidelity. It is ideal if data privacy or on-premise execution is a strict requirement.
    • Azure AI Document Intelligence : An enterprise-grade cloud service featuring prebuilt models for financial and document layouts. It excels at bounding boxes, table cell extraction, and handling multi-page financial statements at scale. Check capabilities via Microsoft Azure Document Intelligence.
    • Google Document AI : Provides specialized prebuilt and custom extractors that parse forms, invoices, and complex tables directly into structured schema. It integrates smoothly into Google Cloud pipelines. Explore options on the Google Cloud Document AI Platform.
    • Nanonets : An AI-driven extraction platform that uses machine learning to read complex financial layouts and tables without needing rigid, fragile template setups. It features a human-in-the-loop review workflow for high-stakes verification.

    Quick Comparison Matrix

    ToolBest ForOutput FormatDeployment Options
    LlamaParseFinancial RAG & complex tablesMarkdown, JSON, structured textCloud, Single-tenant, On-prem
    DoclingOpen-source & local privacyMarkdown, JSON, HTMLLocal / Self-hosted
    Azure Document IntelligenceEnterprise Azure ecosystemsJSON, Structured dataCloud, Container
    Google Document AIGCP-native pipelinesJSON, Custom schemaCloud (GCP)
    NanonetsNo-code UI with validation workflowsJSON, CSV, ExcelCloud, VPC, On-prem

    To help narrow down the ideal choice, tell me:

    • Are you processing millions of pages (high-volume batch) or ad-hoc reports?
    • Do you need to run this locally for data privacy , or is a cloud API fine?
    • Are the reports standardized (like SEC filings) or wildly varied in layout?

    ChatGPTNov 18, 2026

    1. 1Azure AI Document Intelligence
    2. 2Google Cloud Document AI
    3. 3Amazon Textract
    4. 4ABBYY Vantage
    5. 5Docsumo
    6. 6LlamaParse
    7. 7ABBYY FineReader PDF
    8. 8Tabula

    For extracting tables from PDF financial reports, I recommend looking at document AI platforms (Intelligent Document Processing or IDP) rather than relying on traditional OCR software.

    The important distinction is that recognizing text is relatively easy; accurately reconstructing financial tables is much harder. A good solution needs to preserve:

    • Rows and columns, including merged cells.
    • Financial figures, decimal places, negatives and parentheses.
    • Multi-level column headers.
    • Tables that continue across multiple pages.
    • Footnotes, subtotals and grand totals.
    • Relationships between financial labels and their corresponding values.

    Here are the strongest options I'd shortlist as of November 2026.

    1. Recommended document AI software

    • Analyze Forms and Documents with Azure AI Document Intelligence | CloudLearn

    1. Azure AI Document Intelligence

    (Best overall starting point)

    Microsoft

    • Extracts tables, rows, columns, cell positions and merged-cell information.
    • Supports scanned and digitally generated PDFs.
    • Returns structured JSON suitable for Excel, databases and downstream processing.
    • Offers custom models for specialized document formats.

    Best for: Financial reports with complex layouts and organizations wanting a customizable extraction pipeline.

    Limitation: Tables spanning pages may need additional processing to combine them into one continuous table.

    Sources:

    Microsoft product website


    • Google Cloud Document AI with API | Generative AI Projects | Codersarts AI

    2. Google Cloud Document AI

    (Best for structured extraction APIs)

    Google

    • Extracts tables, key-value pairs and document structure.
    • Provides pretrained and custom document processors.
    • Supports OCR and multilingual document processing.
    • Integrates with Google Cloud data pipelines.

    Best for: Large collections of financial PDFs where you want to automate extraction into structured datasets.

    Limitation: Tables continuing across multiple pages may be returned separately and require post-processing.

    Sources:

    Google Document AI website


    • Amazon Textract becomes PCI DSS certified, and retrieves even more data from tables and forms | Artificial Intelligence

    3. Amazon Textract

    (Best for cell-level table extraction)

    Amazon Web Services

    • Detects individual table cells and their relationships.
    • Recognizes merged cells, column headers, titles and footers.
    • Supports structured and semi-structured tables.
    • Returns detailed machine-readable results through its API.

    Best for: Financial statements, balance sheets and reports where preserving table structure is critical.

    Limitation: Multi-page table reconstruction generally requires additional development.

    Sources:

    Amazon Textract website


    • What is ABBYY Vantage? | ABBYY

    4. ABBYY Vantage

    (Best for enterprise document processing)

    ABBYY

    • Intelligent document processing with configurable extraction workflows.
    • Supports OCR, classification and structured data extraction.
    • Offers validation and human review capabilities.
    • Suitable for recurring business document workflows.

    Best for: Financial institutions processing large volumes of reports with quality-control requirements.

    Limitation: More involved implementation and enterprise-oriented deployment.

    Source:

    ABBYY Vantage website


    • Invoice Processing Software Comparison: Top 5 Tools for 2025

    5. Docsumo

    (Best for finance-focused extraction)

    Docsumo

    • Designed for financial and business documents.
    • Extracts tables, fields and line items.
    • Supports scanned and digital documents.
    • Provides validation and review workflows.

    Best for: Financial reports, statements and documents that need to become usable Excel or database records.

    Limitation: Confirm that its table extraction handles your particular financial-report layouts before committing.

    Sources:

    Docsumo website


    • Super Easy way to Parse PDFs ✅🔥. Introducing LlamaParse from LlamaIndex | by Sudarshan Koirala | Medium

    6. LlamaParse

    (Best for AI-assisted PDF parsing)

    LlamaIndex

    • Layout-aware document parsing.
    • Can produce structured outputs, including Markdown and JSON.
    • Useful for extracting tables into AI applications and analytical workflows.
    • Supports configurable parsing approaches.

    Best for: Financial-report analysis, document question-answering and extracting tables for further AI processing.

    Limitation: You may need additional work to validate financial values and normalize tables.

    Sources:

    LlamaParse website

    2. Quick comparison

    SoftwareTable extractionScanned PDFsCoding required
    Azure Document IntelligenceExcellentYesUsually
    Google Document AIExcellentYesUsually
    Amazon TextractExcellentYesYes
    ABBYY VantageVery goodYesLow-code
    DocsumoVery goodYesLow-code
    LlamaParseVery goodYesUsually

    These are qualitative assessments of capabilities, not independently verified accuracy rankings. Actual performance depends on PDF quality and table complexity.

    3. Other tools worth considering

    Depending on your workflow, two additional options are worth testing:

    • ABBYY FineReader PDF: Particularly useful if you want desktop software to convert PDFs into Excel without building an API pipeline. Website.
    • Tabula: Free, open-source software for extracting tables from digitally generated PDFs. It is less suitable for scanned documents. Website.

    4. What I would test before choosing

    Financial reports have some unusual table-extraction challenges. I would create a small benchmark using 10–20 representative reports.

    Suggested evaluation criteria

    of criteria evaluated [Reset]

    One particularly important test: Extract a balance sheet and income statement, then compare the extracted numbers against the original PDF. A tool can achieve excellent OCR accuracy but still put a value in the wrong financial column.

    5. My recommendation

    I'd narrow the initial shortlist to three:

    1. Azure AI Document Intelligence – best starting point for a customizable extraction solution.
    2. Amazon Textract – worth testing for detailed cell-level table reconstruction.
    3. Docsumo – worth evaluating if you want a more ready-made financial document processing workflow.

    I would not select a winner without testing actual annual reports, quarterly filings and financial statements.

    6. A few questions to narrow down the best solution

    The right choice depends considerably on how you intend to use the extracted data.

    1. What types of financial reports are you processing?

    • Annual reports (10-K, 20-F)
    • Quarterly reports (10-Q)
    • Balance sheets and income statements
    • Bank statements
    • Investment research reports
    • A mixture of these

    2. What output do you need?

    • Excel spreadsheets
    • CSV files
    • JSON for database ingestion
    • Structured tables for an AI/LLM application

    3. How do you want to use the software?

    • No-code software with a user interface
    • API integration into an existing application
    • Python-based automated extraction pipeline
    • Enterprise solution with validation and human review

    4. Approximately how many PDF pages do you process monthly?

    • Under 1,000
    • 1,000–10,000
    • 10,000–100,000
    • More than 100,000

    With these answers, I can narrow this down to the three most appropriate solutions and recommend a practical architecture for extracting financial tables accurately, including handling multi-page tables and exporting clean Excel data.

Document AI for tabular data: common questions

Which brands does AI recommend for Document AI for tabular data?
AI most often recommends Google Cloud Document AI (84%), Azure AI Document Intelligence (74%), Amazon Textract (55%), LlamaParse (38%) and Nanonets (38%) for Document AI for tabular data in AI Document Parsing & Extraction APIs.
What questions do buyers ask AI about Document AI for tabular data?
Common buyer questions include "I cannot extract data from tables in PDF financial reports. Who offers specialized document AI for tabular data?", "I need a document AI tool that can accurately pull tables out of PDF financial reports. What are my best options?", "I need a solution for extracting tables from PDF financial reports. Can you recommend document AI software that handles tabular data well?".
For agents: this page as Markdown, parse.gl/topics/document-ai-for-tabular-data.md.

This page

Reuse: Figures and tables on this page are free to republish with attribution: a link to https://parse.gl/topics/document-ai-for-tabular-data beside them. A standard followed link (dofollow) is preferred.

Read and search

Every brand's share on this topic, with 95% ranges: GET agent.parse.gl/v1/tools/query_metrics?args=%7B%22measures%22%3A%5B%22share_named%22%5D%2C%22by%22%3A%5B%22brand%22%5D%2C%22filter%22%3A%7B%22topic%22%3A%5B%22top_c6333cb1-03c7-4f75-8120-ca6745774ffb%22%5D%7D%7D
Search the index: GET parse.gl/api/public/search?q=<words>

Connect

Connect an assistant: agent.parse.gl/mcp

An agent can save work before signup: POST https://agent.parse.gl/v1/workspaces

What it costs

No account: Public figures and up to 5 examples per list, about 120 brands a day per address.
Free account: $5 in credit, your brand's first check (20 buyer questions) and a free weekly check.
After your first top-up: Exports and deeper search results. The first top-up starts at $5.
Live checks: $0.01 per question per engine ($0.05 on Gemini and Claude).
Full public answers and pages: 25 free per day, then $0.005 each.
Public answers older than 90 days: $0.001 each the first time your workspace counts or lists one, then yours.