PixLab Vision PlatformBrowser Workspace

Understand, query, and edit documents in Vision Workspace

A browser-based environment for opening visual and document content, extracting text with OCR, asking natural-language questions, editing the result, and organizing structured information in Smart Tables.

OCR & Document Parsing Document Query PDF & Text Editor Smart Tables Vision-Language Models
Open Vision Workspace 日本語で見る · View in Japanese

Need the document intelligence to continue into a controlled desktop task? Connect the workflow with Think-Act.

PixLab Vision Workspace home screen with image and document upload panels, model selection, reasoning controls, and navigation for Document Query and OCR, PDF and Text Editor, and Smart Tables
The live Workspace groups visual chat, document uploads, OCR, editing, and tables in one browser interface.
Four connected areas

Upload content, understand it, then organize and edit

Vision Workspace organizes its tools into four connected areas. Choose the area that fits the task, or move between them as your work develops.

Workspace

Your central session view. Upload images, PDFs, or documents, organize open files, and start any document task from one place.

Launch Workspace

Document Query & OCR

Extract text from images, scans, and PDFs using OCR. Then ask natural-language questions to surface specific data, summaries, or field values.

Related Image Query API

PDF & Text Editor

Edit extracted text, write in Markdown, and refine AI-generated content. The editor keeps source documents and editable output in the same session.

Try the editor

Smart Tables

Move extracted fields, parsed values, and structured data into a spreadsheet-like table for review, formatting, or export without leaving the browser.

Related LLM Parse API
See all four areas in Workspace
Workspace interface

Document work stays inside one browser tab

Vision Workspace keeps your files and document tools in one browser environment. Move between Workspace, Document Query & OCR, PDF & Text Editor, and Smart Tables without switching applications.

  • Upload supported images and documents

    The current interface accepts JPG, PNG, WEBP, and BMP images, plus PDF, Excel, Word, and text documents.

  • Review extracted content before using it

    Use the query and editing areas to inspect OCR output, parsed fields, and generated text before copying or exporting a result.

  • Free and account options

    The live Workspace currently presents free and premium model choices alongside PixLab account sign-in. Check vision.pixlab.io ↗ for current access requirements and usage limits.

Start a document task
PixLab Vision Workspace interface with visual chat controls, image and document uploads, model selection, and navigation to its four work areas

Source

PDF · Image · Scan

Process

OCR · Query · Edit

Output

Text · Table · JSON

The Workspace home screen provides visual chat controls, supported image and document uploads, and navigation to Document Query and OCR, PDF and Text Editor, and Smart Tables.
Document intelligence workflow

From raw document to structured result

Vision Workspace connects OCR, document understanding, querying, editing, and structured output as one practical sequence. You can stop after any stage or continue into a developer or agentic workflow.

Vision Workspace Document Query and OCR screen with a document upload area, supported-format note, and document library controls
Document Query and OCR provides a dedicated upload area for supported PDFs, Word and Excel documents, images, and text files, with controls for browsing the document library.
Run a document workflow Query API reference
  1. 01

    Input

    PDF · JPG · PNG · WEBP · BMP

    Upload or open the content

    Open a PDF, image, or scan in the Workspace. No conversion step required. Your document appears in the viewer ready for processing.

  2. 02

    Extraction

    OCR API

    Extract text with OCR

    Optical character recognition converts images, scans, and PDF content into readable, searchable text. The extracted result appears alongside the source document for comparison.

  3. 03

    Query

    Query API

    Ask questions in natural language

    Use the document query interface to surface specific values, field data, summaries, or comparisons from the uploaded content. Review the answer against the source before using it downstream.

  4. 04

    Parsing

    LLM Parse API

    Parse into structured data

    Move beyond plain text extraction. Structured parsing converts document content into JSON, Markdown, or organized fields — ready for downstream use, APIs, or export to Smart Tables.

  5. 05

    Review & edit

    Review, edit, and use the result

    Refine extracted or generated content in the text editor. Organize field data in Smart Tables. Continue manually, use related PixLab APIs in an application, or hand the structured output to a controlled agentic workflow in Think-Act.

Vision Workspace PDF and Text Editor with selected text, an open formatting menu, and fullscreen, export, copy, AI Assist, and clear actions

Bring in

Extracted text

Refine

Edit · Transform

Finish

Copy · Export

The PDF and Text Editor combines text formatting with fullscreen, export, copy, AI Assist, and clear controls.
PDF & Text Editor

Move from an answer to an editable draft

The editor closes the gap between understanding a document and producing usable text. Start a draft, bring across extracted or generated content, edit it manually, then use targeted AI assistance only where it helps.

Generate

Draft new text from an instruction.

Transform

Rewrite, summarize or change tone on selected text.

Review

Keep manual editing in the loop before export.

Export

Move the finished draft into the next tool or workflow.

Edit in Vision Workspace
Smart Tables

When the answer belongs in rows and columns, treat it like data

Smart Tables give extracted or manually entered information a spreadsheet-style surface for review, cleanup, formulas, and export without moving the work into another tool.

Verify extracted values
Apply table operations
Export reviewed data

Spreadsheet operations

Use familiar functions while cleaning or shaping extracted data.

AVG ROUND MIN MAX CONCAT LEFT

Export choices

Use the Workspace export control after reviewing and preparing the table.

Review the current export options in the live Workspace and choose the format available for your next step.

A useful pattern: extract fields from a report or invoice, verify the values against the source, arrange them in a table, apply formulas where needed, then export the reviewed dataset.

Useful document workflows

Practical tools for teams that work with documents

Use Vision Workspace to review contracts, reconcile financial records, process forms, or explore a document task before building an integration.

Legal teams

Extract key terms, clauses, and dates from contracts and scanned agreements. Use document query to locate specific provisions without reading the entire document.

Clause extraction Contract query Date parsing

Finance & accounting

Parse invoices, receipts, and financial statements into structured field data. Move extracted values directly into Smart Tables for review and reconciliation.

Invoice parsing Smart Tables export Reconciliation

HR & operations

Extract structured data from submitted forms and scanned documents. Use the editor to reformat and prepare content for records or downstream systems.

Form extraction CV parsing Records prep

Developers & analysts

Use the browser Workspace to explore document tasks before building an integration. PixLab provides separate developer APIs for related OCR, image-query, parsing, and document-processing workflows.

API prototyping Logic verification REST integration
Try it with your document
Access & accounts

Start free. Scale when you need to

The Workspace currently presents free and premium model choices, account sign-in, a document library, and links to the broader PixLab account environment. For current access requirements and usage limits, check the live Workspace; pricing and availability can change.

Check current access PixLab pricing

Account-based access

Vision Workspace exposes sign-in and account creation. The connected PixLab account environment provides API keys, usage data, and support controls.

Free & premium model tiers

The Workspace labels both free and premium model areas. Check the live product for current model availability, usage limits, and plan requirements.

AWS S3 integration

The PixLab account environment includes an AWS S3 integration area for storage workflows. Check current documentation for scope and availability.

Sensitive-document workloads

For regulated use cases, confirm current data handling and contractual requirements through PixLab's published policies or support channels before production use.

Developer and document ecosystem

Build document intelligence into your application

PixLab provides separate REST APIs for related developer workflows, including OCR, image query, structured document parsing, LLM tools, and identity-document scanning. Choose the endpoint whose documented inputs and outputs match your application.

Browse all Vision APIs API portal
Server-side JavaScript · async parse + poll 
const API_KEY = process.env.PIXLAB_API_KEY;
const SUBMIT_URL = 'https://api.pixlab.io/llmparse';
const JOB_URL = 'https://api.pixlab.io/job';

async function parseDocumentForLlm() {
  const submitResponse = await fetch(SUBMIT_URL, {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json'
    },
    body: JSON.stringify({
      key: API_KEY,
      url: 'https://example.com/report.pdf',
      format: 'md',
      extension: 'pdf'
    })
  });

  if (!submitResponse.ok) {
    throw new Error(
      `LLM Parse request failed: ${submitResponse.status}`
    );
  }

  const job = await submitResponse.json();

  if (!job.rc) {
    throw new Error(
      job.err ||
      job.error ||
      'LLM parse job was not accepted'
    );
  }

  while (true) {
    const statusResponse = await fetch(
      `${JOB_URL}/${job.jobId}?key=${encodeURIComponent(API_KEY)}`
    );

    if (!statusResponse.ok) {
      throw new Error(
        `Job status request failed: ${statusResponse.status}`
      );
    }

    const status = await statusResponse.json();

    if (status.status === 'completed') {
      console.log(status.result.format);
      console.log(status.result.data);
      return status.result;
    }

    if (status.status === 'failed') {
      throw new Error(
        status.result?.error ||
        'Document parsing failed'
      );
    }

    await new Promise(resolve =>
      setTimeout(resolve, 2000)
    );
  }
}

parseDocumentForLlm().catch(console.error);

/chat

Document chat

/llm-tools

Tool calling

REST API

All endpoints
Workspace → controlled desktop workflow

Vision Workspace understands. Think-Act acts

Vision Workspace is where you inspect, query, extract, and edit document intelligence in the browser. Think-Act is PixLab's Windows desktop agent that extends that intelligence into controlled, bounded desktop workflows with explicit approval gates and reviewable results.

Vision Workspace

  • Browser-based, no install
  • OCR, document query, editing
  • Smart Tables and export
  • Free model option

Think-Act Desktop

  • Windows desktop agent
  • Multi-step agentic workflows
  • Approval controls and logs
  • Reusable local skills
FAQ

Common questions about Vision Workspace

Find out what Vision Workspace supports, how to get started, and how it relates to Think-Act and the PixLab APIs. Can't find the answer you need? Contact support.

Visit Vision Workspace Think-Act FAQ
What is PixLab Vision Workspace?

Vision Workspace is PixLab's browser-based environment for working with visual and document content. It includes OCR and document parsing, natural-language document queries, PDF and text editing, and Smart Tables for organizing extracted data — all accessible without installation.

What document formats does Vision Workspace support?

The current Workspace interface lists JPG, PNG, WEBP, and BMP for images, and PDF, Excel, Word, and text files for documents. Check the live Workspace ↗ for the current complete list of supported inputs.

Can I use Vision Workspace without signing up?

Vision Workspace can be opened without installation and currently presents a free model option alongside account sign-in and creation controls. Whether a particular task requires an account, and the applicable usage limits, can change; check the live Workspace ↗ for current access details.

How does document query work in Vision Workspace?

In the Document Query & OCR area, upload a supported file and use natural-language prompts to query, summarize, and explore its content. PixLab's separate /query developer endpoint is documented for asking questions about a submitted image, so its behavior should not be assumed identical to the browser workflow.

What is the difference between Vision Workspace and Think-Act?

Vision Workspace ↗ is a browser-based environment for interactive document work: OCR, extraction, querying, editing, and organizing content. Think-Act is PixLab's Windows desktop AI agent for controlled, reviewable multi-step workflows that can span tools and applications. The two products are connected but serve different modes of work.

Is Vision Workspace connected to PixLab's developer APIs?

PixLab provides separate developer APIs for related workflows: /ocr for supported images or video frames, /query for image questions and answers, /llm-parse for supported documents, and DocScan for supported identity documents. These APIs can be used independently; do not assume an individual endpoint behaves identically to Vision Workspace. Start with the VLM API reference hub or explore LLM and parsing APIs.

Start with Vision Workspace

Open a document and see what the Workspace finds

Upload an image or PDF, run OCR, ask a question about its content, and review the result. Because the Workspace runs in the browser, there is nothing to install or configure. Start with one practical document task and build from there.

日本語で見る (View in Japanese)
Launch Vision Workspace Browser document tools · Free Download Think-Act Windows desktop agent · Free
Platform Vision Platform overview Agents AI Agent workflows Documents LLM & Parsing APIs Developers Vision APIs