> ## Documentation Index
> Fetch the complete documentation index at: https://help.abacusdocs.com/llms.txt
> Use this file to discover all available pages before exploring further.

# What is Abacus Docs?

> Abacus Docs extracts text and structured fields from documents, then lets you batch, share, and ask questions.

Abacus Docs is a document intelligence product at [abacusdocs.com](https://abacusdocs.com). You upload PDFs and images. The platform transcribes the pages, can extract named fields against a schema you design, and keeps the results in a library you can search, share, and query in chat.

It is built for teams that already live in documents — invoices, contracts, filings, forms, shipping papers — and need the contents in a grid or a conversation, not another pile of files.

## The pieces you will use

<Columns cols={2}>
  <Card title="Documents" icon="file-text" href="/guides/uploading-documents">
    A file you upload lands in **My Documents**. Opening it shows the original next to transcription, sections, and any schema fields.
  </Card>

  <Card title="Schemas" icon="table" href="/schemas/overview">
    An output schema (listed as **Schemas** in the sidebar) names the fields to extract — invoice number, total, dates, tables. You author them in the schema builder.
  </Card>

  <Card title="Batches" icon="layer-group" href="/guides/batch-processing">
    A batch job runs the same schema across many files. The sidebar label is **Batch Processing**; the page title is **Intelligent Document Processing**.
  </Card>

  <Card title="Chat" icon="comments" href="/guides/chat-and-analysis">
    Two chats exist: **Chat** in the sidebar (over threads and attached batches) and the **Chat** tab on a single document (Deep Reader).
  </Card>
</Columns>

## How a file is processed

<Steps>
  <Step title="Upload">
    Drop a PDF or image on **My Documents**, or include files in a **New Batch Job**. Each file can be up to **64 MB**. A batch can hold up to **1,024** files.
  </Step>

  <Step title="Extraction">
    Abacus Docs reads every page (typed text, scans, handwriting) and produces a transcription. If you applied a schema, it also fills those fields with values, confidence, and a source page.
  </Step>

  <Step title="Review and use">
    Open the [document viewer](/concepts/document-processing), edit cells in a [batch grid](/guides/exporting-results), [share](/sharing/public-links) a read-only link, or [ask a question](/guides/chat-and-analysis).
  </Step>
</Steps>

## Credits

Processing spends **credits** from your organization's balance. New individual accounts receive **1,000 credits** with no card. See [Credits and billing](/account/credits).

<Note>
  Your uploaded files and the outputs we generate for you are not used to train AI models. You keep intellectual property ownership. Details are on [Data privacy](/security/data-privacy).
</Note>