LLMWhisperer

LLMWhisperer is a document-to-text extraction API from Unstract (Zipstack) that turns complex PDFs, scanned documents, and images into clean, layout-preserving text ready for large language models. It exposes an asynchronous REST API (v2) - submit a document to /whisper, poll /whisper-status, then retrieve the extracted text via /whisper-retrieve - plus line-level highlight coordinates and webhook callbacks. Authentication is via the unstract-key header.

5 APIs 0 Features

AILLMDocument ExtractionOCRText Extraction

APIs

LLMWhisperer Whisper Extraction API

Submits a document (PDF, image, or URL) to POST /whisper for asynchronous, layout-preserving text extraction across native_text, low_cost, high_quality, form, and table modes. R...

LLMWhisperer Whisper Status API

GET /whisper-status returns the processing state (accepted, processing, processed, error, retrieved) for a whisper_hash, with per-page execution detail.

LLMWhisperer Whisper Retrieve API

GET /whisper-retrieve returns the extracted result_text plus optional confidence_metadata for a processed whisper_hash. Results can be retrieved once.

LLMWhisperer Highlights API

GET /highlights returns per-line bounding-box coordinates (base_y, height, page, page_height) for the requested lines so callers can highlight extracted text in the source docum...

LLMWhisperer Webhooks API

Register and manage webhook callbacks via /whisper-manage-callback (POST/GET/PUT/DELETE). Submit a document with use_webhook to have the extracted result delivered to your endpo...

Resources

👥

GitHubOrganization

Sources

aid: llmwhisperer
url: https://raw.githubusercontent.com/api-evangelist/llmwhisperer/refs/heads/main/apis.yml
name: LLMWhisperer
kind: company
description: LLMWhisperer is a document-to-text extraction API from Unstract (Zipstack)
  that turns complex PDFs, scanned documents, and images into clean, layout-preserving
  text ready for large language models. It exposes an asynchronous REST API (v2) -
  submit a document to /whisper, poll /whisper-status, then retrieve the extracted
  text via /whisper-retrieve - plus line-level highlight coordinates and webhook
  callbacks. Authentication is via the unstract-key header.
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
tags:
- AI
- LLM
- Document Extraction
- OCR
- Text Extraction
created: '2026-06-20'
modified: '2026-06-20'
specificationVersion: '0.19'
apis:
- aid: llmwhisperer:whisper-extraction-api
  name: LLMWhisperer Whisper Extraction API
  tags:
  - Extraction
  - OCR
  - Layout Preserving
  image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
  humanURL: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_text_extraction_api/
  baseURL: https://llmwhisperer-api.us-central.unstract.com/api/v2
  properties:
  - url: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_text_extraction_api/
    type: Documentation
  - url: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_apis_intro/
    type: APIReference
  - url: openapi/llmwhisperer-openapi.yml
    type: OpenAPI
  - url: collections/llmwhisperer.postman_collection.json
    type: PostmanCollection
  - url: collections/llmwhisperer.opencollection.json
    type: OpenCollection
  description: Submits a document (PDF, image, or URL) to POST /whisper for asynchronous,
    layout-preserving text extraction across native_text, low_cost, high_quality,
    form, and table modes. Returns a 202 with a whisper_hash used to track and
    retrieve the job.
- aid: llmwhisperer:whisper-status-api
  name: LLMWhisperer Whisper Status API
  tags:
  - Status
  - Async
  - Polling
  image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
  humanURL: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_text_extraction_status_api/
  baseURL: https://llmwhisperer-api.us-central.unstract.com/api/v2
  properties:
  - url: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_text_extraction_status_api/
    type: Documentation
  - url: openapi/llmwhisperer-openapi.yml
    type: OpenAPI
  - url: collections/llmwhisperer.postman_collection.json
    type: PostmanCollection
  - url: collections/llmwhisperer.opencollection.json
    type: OpenCollection
  description: GET /whisper-status returns the processing state (accepted, processing,
    processed, error, retrieved) for a whisper_hash, with per-page execution detail.
- aid: llmwhisperer:whisper-retrieve-api
  name: LLMWhisperer Whisper Retrieve API
  tags:
  - Retrieve
  - Results
  - Text
  image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
  humanURL: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_text_extraction_retrieve_api/
  baseURL: https://llmwhisperer-api.us-central.unstract.com/api/v2
  properties:
  - url: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_text_extraction_retrieve_api/
    type: Documentation
  - url: openapi/llmwhisperer-openapi.yml
    type: OpenAPI
  - url: collections/llmwhisperer.postman_collection.json
    type: PostmanCollection
  - url: collections/llmwhisperer.opencollection.json
    type: OpenCollection
  description: GET /whisper-retrieve returns the extracted result_text plus optional
    confidence_metadata for a processed whisper_hash. Results can be retrieved once.
- aid: llmwhisperer:highlights-api
  name: LLMWhisperer Highlights API
  tags:
  - Highlights
  - Bounding Box
  - Coordinates
  image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
  humanURL: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_highlighting_api/
  baseURL: https://llmwhisperer-api.us-central.unstract.com/api/v2
  properties:
  - url: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_highlighting_api/
    type: Documentation
  - url: openapi/llmwhisperer-openapi.yml
    type: OpenAPI
  - url: collections/llmwhisperer.postman_collection.json
    type: PostmanCollection
  - url: collections/llmwhisperer.opencollection.json
    type: OpenCollection
  description: GET /highlights returns per-line bounding-box coordinates (base_y,
    height, page, page_height) for the requested lines so callers can highlight
    extracted text in the source document.
- aid: llmwhisperer:webhooks-api
  name: LLMWhisperer Webhooks API
  tags:
  - Webhooks
  - Callbacks
  - Async
  image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
  humanURL: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_webhooks_manage/
  baseURL: https://llmwhisperer-api.us-central.unstract.com/api/v2
  properties:
  - url: https://docs.unstract.com/llmwhisperer/llm_whisperer/apis/llm_whisperer_webhooks_manage/
    type: Documentation
  - url: openapi/llmwhisperer-openapi.yml
    type: OpenAPI
  - url: review.yml
    type: Review
  - url: collections/llmwhisperer.postman_collection.json
    type: PostmanCollection
  - url: collections/llmwhisperer.opencollection.json
    type: OpenCollection
  description: Register and manage webhook callbacks via /whisper-manage-callback
    (POST/GET/PUT/DELETE). Submit a document with use_webhook to have the extracted
    result delivered to your endpoint instead of polling.
common:
- type: GitHubOrganization
  url: https://github.com/Zipstack
- type: LinkedIn
  url: https://www.linkedin.com/company/unstract
- type: Website
  url: https://unstract.com/llmwhisperer/
- type: Documentation
  url: https://docs.unstract.com/llmwhisperer/
- type: Plans
  url: plans/llmwhisperer-plans-pricing.yml
- type: RateLimits
  url: rate-limits/llmwhisperer-rate-limits.yml
- type: FinOps
  url: finops/llmwhisperer-finops.yml
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com