---
title: "Create Empty Document"
method: POST
path: "/api/v2/ml-scribe/open-api/sessions/{session_id}/documents"
tags: ["Heidi API"]
---

# Create Empty Document

`POST /api/v2/ml-scribe/open-api/sessions/{session_id}/documents`

create an empty document for the session. This document is a placeholder for the AI generation in next step.
Args:
    session_id (str)

Returns:
    Document

## Path parameters

- `session_id` string, required

## Request body

- OpenAPIGenerateDocumentContentRequestV2
  - `document_tab_type` 'DOCUMENT' | 'DICTATION' | 'PDF'
  - `generation_method` 'TEMPLATE' | 'THIRD_PARTY_TEMPLATE', required
  - `template_id` string, nullable
  - `voice_style` 'BRIEF' | 'GOLDILOCKS' | 'DETAILED' | 'MY_VOICE' | 'SUPER_DETAILED' — was called writing style
  - `custom_style` CustomStyle
    - `level_of_detail` 'LOW' | 'MEDIUM' | 'HIGH', required
    - `bullet_points_preference` boolean — Whether to use bullet points in the note
    - `patient_quotes_preference` boolean — Whether to include patient quotes in the note
    - `abbreviations_preference` boolean — Whether to use abbreviations in the note
  - `ai_command_id` string, nullable
  - `ai_command_text` string, nullable
  - `dictation_cleanup_mode` 'SMART' | 'WORD_FOR_WORD'
  - `brain` 'LEFT' | 'RIGHT' — Brain enum for different AI model configurations. Each brain represents a different AI model configuration with specific characteristics: - LEFT: Fast and simple model - RIGHT: Creative and artistic model
  - `index` integer, nullable
  - `content_type` 'MARKDOWN'
  - `live_dictation_text` string, nullable
  - `is_personalization_applied` boolean, nullable
  - `is_not_regeneration` boolean, nullable
  - `task_id` string, nullable

## Response `200`

Successful Response

- SessionDocument
  - `id` string, required
  - `session_id` string, required
  - `creator_kinde_id` string, nullable
  - `document_tab_type` 'DOCUMENT' | 'DICTATION' | 'PDF'
  - `index` integer, nullable
  - `name` string, nullable
  - `template_id` string, nullable
  - `ai_command_id` string, nullable
  - `ai_command_text` string, nullable
  - `dictation_cleanup_mode` 'SMART' | 'WORD_FOR_WORD'
  - `question` string, nullable
  - `original_content` string, nullable
  - `content` string, nullable
  - `updated_dictation` string, nullable
  - `created_at` string, date-time
  - `updated_at` string, date-time
  - `deleted_at` string, date-time, nullable
  - `generation_type` 'TEMPLATE'
  - `brain` 'LEFT' | 'RIGHT' — Brain enum for different AI model configurations. Each brain represents a different AI model configuration with specific characteristics: - LEFT: Fast and simple model - RIGHT: Creative and artistic model
  - `feedback` Feedback
    - `rating` 1 | 2 | 3 | 4 | 5, required
    - `feedback` FeedbackRatingReason[]
      - `reason` string, required
    - `detail` string
    - `updated` string, date-time
  - `voice_style` 'BRIEF' | 'GOLDILOCKS' | 'DETAILED' | 'MY_VOICE' | 'SUPER_DETAILED' — was called writing style
  - `custom_style` CustomStyle
    - `level_of_detail` 'LOW' | 'MEDIUM' | 'HIGH', required
    - `bullet_points_preference` boolean — Whether to use bullet points in the note
    - `patient_quotes_preference` boolean — Whether to include patient quotes in the note
    - `abbreviations_preference` boolean — Whether to use abbreviations in the note
  - `content_type` 'MARKDOWN'
  - `generation_time` string, nullable — successful note generation time string in utc timezone, iso8601 format
  - `model` union
    - 'A' | 'B' | 'C' | 'D' — LLM model names used for note generation.
    - LLMModels[]
  - `is_personalization_applied` boolean, nullable
  - `has_deducted_quota` boolean — Tracks whether quota has been deducted for this document. Prevents multiple quota deductions on regeneration. Resets on template ID change.
  - `task_id` string, nullable

## Other responses

- `400` — Bad Request
- `422` — Validation Error

---

[API](https://skmtc.net/heidihealth/apis/heidi-api-documentation.md) · [All operations](https://skmtc.net/heidihealth/apis/heidi-api-documentation/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/heidihealth/heidi-api-documentation/versions/39b46a9d8637/schema)