# Nanonets integration

> Document AI and OCR platform: upload documents to models and workflows, extract fields and tables, train custom models, review and approve results.

- Authentication: apiKey

## Tools (50)

- **Add Field Value**: Manually add a field value to a document page.
- **Add Table**: Manually add a table, optionally with cells, to a document page.
- **Add Table Cell**: Add a cell to an existing table on a document page.
- **Approve File**: Mark an uploaded file as approved (verified), ready for export.
- **Assign Files To Member**: Assign uploaded files to a team member for review or approval.
- **Classify Image From File**: Upload a local image to an image classification model and get label probabilities.
- **Classify Image From URL**: Send image URLs to an image classification model and get label probabilities.
- **Create OCR Model**: Create a new OCR model with a list of field categories to extract; returns the model_id used by all other v2 calls.
- **Create Workflow**: Create a workflow of a given type; instant learning workflows need fields set afterward.
- **Delete Document**: Permanently delete a processed document and its data from a workflow.
- **Delete Field Value**: Delete a field value from a document page.
- **Delete File**: Delete an uploaded file and its results from a model.
- **Delete Table**: Delete a table from a document page.
- **Delete Table Cell**: Delete a cell from a table on a document page.
- **Delete Workflow Field**: Remove one field or table header from a workflow.
- **Execute External Integration Query**: Run a SQL query against a linked database integration, including updates.
- **Get All Prediction Files**: List the files predicted by an OCR model within a day range, split into moderated (approved) and unmoderated.
- **Get Document**: Fetch full extraction results for a document; use it to collect async results.
- **Get Document Page**: Fetch extraction results for one page of a document.
- **Get Model State**: Fetch an OCR model's training state and status; state 5 means it can serve predictions.
- **Get Prediction File By File ID**: Fetch the prediction results of an entire uploaded file (all pages) by request_file_id; use it to collect async results.
- **Get Prediction File By Page ID**: Fetch the prediction, moderated and raw OCR results for one page of an uploaded file.
- **Get Workflow**: Fetch one workflow's configuration: fields, table headers and settings.
- **Get Workflow Types**: List the workflow types usable when creating a workflow (instant learning and pre-trained document types).
- **List External Integrations**: List the external integrations (databases) linked to the account, with ids used by Execute Query.
- **List Workflow Documents**: List documents processed by a workflow, newest first, with status and basic page info.
- **List Workflows**: List all workflows in the account with their fields, table headers and settings.
- **Predict From File (Async)**: Upload a local file to an OCR model for asynchronous processing; for files over 3 pages. Returns immediately with a request_file_id.
- **Predict From File (Sync)**: Upload a local file to an OCR model and get extracted fields and tables back in the response; for files of 3 pages or fewer.
- **Predict From URL (Async)**: Send a public file URL to an OCR model for asynchronous processing; for files over 3 pages. Returns immediately with a request_file_id.
- **Predict From URL (Sync)**: Send a public file URL to an OCR model and get extracted fields and tables back in the response; for files of 3 pages or fewer.
- **Retry File Exports**: Trigger or retry the configured export for files, for example after a failed export.
- **Retry File Prediction**: Retry prediction for specific files on an existing model.
- **Set Workflow Fields**: Replace all of a workflow's fields and table headers; required to configure instant learning workflows.
- **Train Model**: Start training an OCR model after training images are uploaded, or retrain it after adding more.
- **Unapprove File**: Revert a previously approved file to unapproved.
- **Update AI Agent Guidelines**: Set instructions the model follows when predicting fields and tables; Instant Learning models only.
- **Update Fields In A File**: Edit, add or remove extracted fields on a processed page by sending back its moderated_boxes.
- **Update Field Value**: Manually correct an extracted field value on a document page.
- **Update Table Cell Value**: Manually correct the text of a table cell.
- **Update Workflow Field**: Rename one field or table header of a workflow.
- **Update Workflow Metadata**: Change a workflow's description.
- **Update Workflow Settings**: Enable or disable table capture for a workflow.
- **Upload Document To Workflow**: Upload a file or URL to a workflow for extraction, synchronously or asynchronously.
- **Upload Training Images By File**: Upload a local image with its field annotations as training data for an OCR model.
- **Upload Training Images By URL**: Upload training images by public URL with their field annotations to an OCR model.
- **Verify Document**: Mark every field and table in a document as verified.
- **Verify Field**: Mark a single extracted field value as verified.
- **Verify Table**: Mark a whole table, with all its cells, as verified.
- **Verify Table Cell**: Mark a single table cell as verified.

Connect Nanonets in General Input: https://www.generalinput.com/apps/nanonets