# Predictions from documents

> Predictions from documents - Multiple methods are available for making predictions with Document AI.

This Markdown file sits beside the HTML page at the same path (with a `.md` suffix). It summarizes the topic and lists links for tools and LLM context.

Companion generated at `2026-04-24T16:03:56.604782+00:00` (UTC).

## Primary page

- [Predictions from documents](https://docs.datarobot.com/en/docs/classic-ui/modeling/special-workflows/doc-ai/doc-ai-predictions.html): Full documentation for this topic (HTML).

## Related documentation

- [Classic UI documentation](https://docs.datarobot.com/en/docs/classic-ui/index.html): Linked from this page.
- [Modeling](https://docs.datarobot.com/en/docs/classic-ui/modeling/index.html): Linked from this page.
- [Specialized workflows](https://docs.datarobot.com/en/docs/classic-ui/modeling/special-workflows/index.html): Linked from this page.
- [Document AI](https://docs.datarobot.com/en/docs/classic-ui/modeling/special-workflows/doc-ai/index.html): Linked from this page.
- [UI](https://docs.datarobot.com/en/docs/classic-ui/modeling/analyze-models/predictions/predict.html): Linked from this page.
- [API](https://docs.datarobot.com/en/docs/classic-ui/predictions/realtime/index.html): Linked from this page.
- [Portable Prediction Server](https://docs.datarobot.com/en/docs/classic-ui/predictions/port-pred/index.html): Linked from this page.
- [Portable Batch Predictions](https://docs.datarobot.com/en/docs/classic-ui/predictions/batch/index.html): Linked from this page.

## Documentation content

# Predictions from documents

The following prediction methods are available for Document AI.

| Method | Description |
| --- | --- |
| UI | Upload an archive or dataset file for predictions from the UI. |

For all other types other than deployment batch predictions, you must include PDF documents as base64-encoded strings. The public API client includes a utility function to help with conversion.

| Method | Description |
| --- | --- |
| API | Use scripting code and base64-converted document files to make an API call and get predictions from the deployed model. The output will be a CSV file with prediction results. |
| Portable Prediction Server | Use base64-converted document files for either single-model or multi-model modes. |
| Portable Batch Predictions | Use base64-converted document files for all supported adapters (filesystem, JDBC, AWS S3, Azure Blob, GCS, Snowflake, Synapse). |