Skip to main content

Modes

LlamaParse supports different parsing modes and presets to adapt to your use case.

Cost-effective​

The cost-effective mode is great for everyday documents that include a mix of images, tables, and text. You get reliable formatting without any extra setup.

Under the Hood​

Cost-effective has been designed for more standard documents, with simple layouts, and for huge batch of documents when the output accuracy isn't critical. It allow you to get good overall results while keeping some credits for other tasks.

To use Cost-effective, set the following parameters:

parser = LlamaParse(
  api_key="<you-api-key>",  # See how to get your API key at https://docs.cloud.llamaindex.ai/api_key
  parse_mode="parse_page_with_llm",  # The parsing mode
  high_res_ocr=True,  # Whether to use high resolution OCR (slower but more precise)
  adaptive_long_table=True,  # Adaptive long table. LlamaParse will try to detect long table and adapt the output
  outlined_table_extraction=True,  # Whether to try to extract outlined tables
  output_tables_as_HTML=True,  # Whether to output tables as HTML in the markdown output
)

Agentic​

Agentic mode has been designed for most documents with diagrams and images. It provides a great quality on the vast majority of documents at a reasonable price. It may struggle with complex layouts or charts.

Under the Hood​

Agentic mode performs OCR, image extraction, and identifies structural elements such as tables and headings. It also outputs equations in LaTeX format and converts diagrams into Mermaid syntax for further analysis and visualization.

To use Agentic, set the following parameters:

parser = LlamaParse(
  api_key="<you-api-key>",  # See how to get your API key at https://docs.cloud.llamaindex.ai/api_key
  parse_mode="parse_page_with_agent",  # The parsing mode
  model="openai-gpt-4-1-mini",  # The model to use
  high_res_ocr=True,  # Whether to use high resolution OCR (slower but more precise)
  adaptive_long_table=True,  # Adaptive long table. LlamaParse will try to detect long table and adapt the output
  outlined_table_extraction=True,  # Whether to try to extract outlined tables
  output_tables_as_HTML=True,  # Whether to output tables as HTML in the markdown output
)

Agentic Plus​

Agentic Plus Mode is designed for complex, highly formatted documents—things like scanned financial reports, research papers, or anything with charts, tables, and diagrams. It gives you the most accurate and structured results.

Under the Hood​

Agentic Plus combines all the capabilities of Agentic mode with enhanced processing power and a better model. It's able to handle the most challenging document layouts, providing superior accuracy for complex tables, charts, and multi-column layouts.

To use Agentic Plus, set the following parameters:

parser = LlamaParse(
  api_key="<you-api-key>",  # See how to get your API key at https://docs.cloud.llamaindex.ai/api_key
  parse_mode="parse_page_with_agent",  # The parsing mode
  model="anthropic-sonnet-4.0",  # The model to use
  high_res_ocr=True,  # Whether to use high resolution OCR (slower but more precise)
  adaptive_long_table=True,  # Adaptive long table. LlamaParse will try to detect long table and adapt the output
  outlined_table_extraction=True,  # Whether to try to extract outlined tables
  output_tables_as_HTML=True,  # Whether to output tables as HTML in the markdown output
)

Use Case Specific Presets​

Use-case specific preset are versioned. You can get the last version of the preset by calling them with the -latest flag or directly without flags.

Invoices​

For Invoices and receipts. Set preset="invoice" on API. Current version is invoice-v-1.

Scientific Paper​

For Scientific papers. Set preset="scientific" on API. Current version is scientific-v-1.

Technical Documentation​

For technical documentation with schematics. Set preset="technicalDocumentation" on API. Current version is technicalDocumentation-v-1.

Output schematics in the present in the documentation using the following XML format in the markdown:

<schematic>
    <bbox>0.1,0.1,0.9,0.9</bbox>
    <caption>Caption of the schematic (if present)</caption>
    <schematic_elements>
        <element>
            <type>Type of the element (e.g. resistor, capacitor, etc.)</type>
            <value>Value of the element (e.g. 10k, 100uF, etc.)</value>
            <position>x,y</position>
            <label>Label of the element (if present)</label>
        </element>
        <!-- Repeat for each element in the schematic -->
    </schematic_elements>
</schematic>

Forms​

For forms and questionnaires. Set preset="forms" on API. Current version is forms-v-1.

Output form field in the markdown as a table with the following columns Field Name, Field Type, Field Value.

| Field Name                     | Field Type   | Field Value                              |
| ------------------------------ | ------------ | ---------------------------------------- |
| A checkbox                     | Checkbox     | Unchecked                                |
| Another checkbox               | Checkbox     | Checked                                  |
| Number of Thin                 | Text         | 1                                        |
| Name of tool                   | Text         | LlamaParse                               |
| Date review                    | Date         | 24/06/2024                               |
| Contractor                     | Signature    | John Doe                                 |
| Available                      | Radio        | Yes                                      |