> This page is for version v2026-02-09 (default).
> For other versions, use one of these documentation indexes:
> - v2026-02-09 (default): https://docs.extend.ai/2026-02-09/llms.txt
> - v2025-04-21: https://docs.extend.ai/2025-04-21/llms.txt
> - v2024-12-23: https://docs.extend.ai/2024-12-23/llms.txt

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.extend.ai/llms.txt
> Use this file to discover all available pages before exploring further.
>
> ## API version
> The current API version is `2026-02-09`, served at the site root (no version prefix in URLs).
> If this page URL contains `/2025-04-21/` or `/2024-12-23/`, you are reading an older API version.
> Prefer the current docs at https://docs.extend.ai/llms.txt unless the user explicitly needs that older version.
> Do not treat older-version pages as the source of truth for new integrations.

# Parse Step

> Configure file parsing behavior for your workflow.

The Parse step is the first step in every workflow and runs automatically when files are submitted.
By default, parser settings are inherited from the document processors configured downstream in your workflow.
However, you can override this behavior by configuring custom parser settings directly on the Parse step.

## When to Configure the Parse Step

Configure the Parse step when you need:

* **Explicit control over parsing** - Override the default settings inherited from downstream processors
* **Consistent parsing across all branches** - Ensure the same parse configuration is used regardless of which extraction, classification, or splitting steps run

## Via the API

In a workflow definition the Parse step is `type: "PARSE"` and must come immediately after the trigger. Omit `config` to inherit parser settings from the downstream processors; set `config.parseConfig` to override them for the whole workflow. `parseConfig` takes the same fields as the standalone parse endpoint; see [Parse Configuration Options](/parsing/configuration).

```json
{
  "name": "parse",
  "type": "PARSE",
  "config": {
    "parseConfig": {
      "target": "markdown",
      "chunkingStrategy": { "type": "page" }
    }
  },
  "next": [{ "step": "extract" }]
}
```

See [Parse in the step reference](/workflows/configuring-workflows#parse) for how this fits into a full `steps` array and how to deploy it.

## Via Extend Studio

To configure the Parse step:

1. Click on the Parse step in your workflow diagram
2. Click the "Configure" button

![Parse step card in the workflow diagram with the Configure button](/_fern-img/344157da85ed5b75657f6d3158be8ace8983ecf6fb68ded1ab7ffcbb63c1b4e3.webp)

This opens the Parse step configuration dialog.

### Enabling Custom Settings

By default, "Use custom parser settings" is disabled, meaning the parser will inherit settings from downstream document processors. Toggle this setting on to configure explicit parser behavior.

![Parse step configuration dialog with the "Use custom parser settings" toggle](/_fern-img/14f28b5077ee86d619d978c05e189d1362c7b72bbe451133f4d8f6cc4c37413b.webp)

When custom settings are enabled, you can configure chunking strategy, block options, and advanced options. See the [Parse guide](/parsing/overview#configuration-options) for detailed documentation of all configuration options.

### Builder vs JSON Mode

The configuration panel offers two modes:

* **Builder**: Visual form-based configuration with all options
* **JSON**: View and import parser configuration as JSON

Use JSON mode to copy configurations between workflows or to import settings from the Parse API.

![Parse step configuration in JSON mode](/_fern-img/04443bf22102622c7e14ed6f17392c2ac5812b8699657e7822567b68d1a929a8.webp)

## Viewing Parse Output

After a workflow runs, you can view the parser output in the workflow run review page. The Parse tab appears first in the output tabs and shows:

* Processing time and page count
* Parsed chunks with their content
* Individual blocks with type and position information
* Raw JSON output

![Parse output panel showing blocks highlighted on the document](/_fern-img/aa34611dba7398a5b7af29b6d0f35b7ffd8b4b076238abcd3a5d855d1a1cdb13.webp)

You can switch between different views:

* **Blocks**: Individual content blocks with spatial highlighting on the document
* **Chunks**: Grouped content chunks
* **Markdown**: Formatted chunk content (when using markdown target)
* **JSON**: Raw parser output

Hovering over blocks in the output panel highlights the corresponding region in the document viewer, and vice versa.

## Relationship to Capabilities

Parser settings can also be configured on individual document processors (extractors, classifiers, splitters). When the Parse step has custom settings enabled:

* The Parse step configuration takes precedence
* All downstream processors use the same parsed content
* This ensures consistent parsing across conditional workflow branches

When custom settings are disabled on the Parse step:

* Parser settings are inferred from downstream processors
* Each processor may contribute to the final parse configuration

## Best Practices

### Use Custom Settings When:

* Your workflow has multiple branches that should use identical parsing
* You need parsing optimizations (like disabling figure parsing for speed)
* You're processing specific document types that require particular settings

### Inherit Settings When:

* Downstream processors already have the correct parser configuration
* You want processor-specific parsing behavior
* You're still iterating on your extraction schema

## Related

* [Parse guide](/parsing/overview) - Full documentation of parser configuration options
* [Parse API reference](/api-reference/endpoints/parse/create-parse-run) - API documentation (same configuration schema applies)