Cloudmersive OCR API
✓ Official Vendor SpecAI/MLVisionapiKey20 EndpointsREST
For Agents
Extract text and structured data from images, PDFs, and photographs — including business cards, receipts, and forms — using a Cloudmersive Apikey.
Use for: Extract text from a scanned PDF, I want the text from an image with bounding boxes for each word, Recognise the contact details on a business card photo, Get the line-by-line text and coordinates from a receipt scan
Not supported: Does not translate text, summarise documents, or render PDFs from scratch — use for OCR text and layout extraction from images and PDFs only.
The Cloudmersive OCR API converts scanned images and PDF pages into machine-readable text. It covers plain text extraction, word- and line-level outputs with bounding boxes, photograph cleanup that turns mobile phone captures into legible scans, and structured recognition of business cards, receipts, and forms. Authentication is an Apikey header issued from the Cloudmersive dashboard, and the API exposes 20 endpoints across image, PDF, and photo recognition flows.
Install Jentic One Beta
Connect the Cloudmersive OCR API to your agent
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Cloudmersive OCR API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
1
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh2
Step 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
Capabilities
What an agent can do with Cloudmersive OCR API.
Convert an image to plain text via /ocr/image/toText
Return text with bounding box locations at /ocr/image/to/lines-with-location and /ocr/image/to/words-with-location
Extract text from PDF pages via /ocr/pdf/toText
Recognise structured data from business card photos at /ocr/photo/recognize/business-card
Recognise filled-in form data at /ocr/photo/recognize/form and /ocr/photo/recognize/form/advanced
Track long-running PDF OCR jobs via /ocr/pdf/get-job-status
Use Cases
Patterns agents use Cloudmersive OCR API for, with concrete tasks.
★ Scanned Document Digitisation
Convert a backlog of scanned PDFs into searchable text for archival or compliance. POST /ocr/pdf/toText for plain text output, or use /ocr/pdf/to/lines-with-location when downstream systems need bounding boxes for highlight rendering. Long files use the async job pattern via /ocr/pdf/get-job-status.
POST a PDF to /ocr/pdf/toText and write the returned text into the document index keyed by file ID.
Mobile Receipt and Business Card Capture
Turn phone-camera photos of receipts and business cards into structured records. /ocr/photo/recognize/business-card returns parsed name, title, company, phone, and email; /ocr/photo/recognize/form (and the advanced variant) extract filled-in form values. Suitable for expense or CRM import flows.
POST a phone photo to /ocr/photo/recognize/business-card and write the parsed contact into the CRM.
Layout-Preserving Text Extraction
Reconstruct document layout by combining word-level OCR with bounding boxes. /ocr/image/to/words-with-location returns each word with its coordinates so renderers can position text on a canvas, build searchable highlights, or feed downstream NLP that needs spatial context.
POST an image to /ocr/image/to/words-with-location and reconstruct the layout in a frontend overlay using the returned coordinates.
AI Agent OCR Workflows
Let a Jentic-powered agent extract text from an attachment without writing OCR plumbing. The agent searches Jentic for 'extract text from a PDF' or 'OCR a business card', loads the matching Cloudmersive operation, and executes it. The Apikey header is injected from the Jentic vault.
Search Jentic for 'extract text from a scanned PDF', load the Cloudmersive pdfToText operation, and execute it with the source file.
Key Endpoints
20 endpoints — the cloudmersive ocr api converts scanned images and pdf pages into machine-readable text.
METHOD
PATH
DESCRIPTION
POST
/ocr/image/toText
Extract plain text from an image
POST
/ocr/image/to/words-with-location
Extract words with bounding boxes
POST
/ocr/image/to/lines-with-location
Extract lines with bounding boxes
POST
/ocr/pdf/toText
Extract text from a PDF
POST
/ocr/pdf/to/words-with-location
Extract PDF text with bounding boxes
GET
/ocr/pdf/get-job-status
Check async PDF OCR job status
POST
/ocr/photo/recognize/business-card
Parse a business card photo
POST
/ocr/photo/recognize/form/advanced
Extract form field data with advanced detection
POST
/ocr/image/toText
Extract plain text from an image
POST
/ocr/image/to/words-with-location
Extract words with bounding boxes
POST
/ocr/image/to/lines-with-location
Extract lines with bounding boxes
POST
/ocr/pdf/toText
Extract text from a PDF
POST
/ocr/pdf/to/words-with-location
Extract PDF text with bounding boxes
GET
/ocr/pdf/get-job-status
Check async PDF OCR job status
POST
/ocr/photo/recognize/business-card
Parse a business card photo
POST
/ocr/photo/recognize/form/advanced
Extract form field data with advanced detection
Why Jentic?
Three things that make agents converge on Jentic-routed access.
Credential isolation
The Cloudmersive Apikey is stored encrypted in the Jentic vault. Agents receive scoped access tokens — the raw Apikey never enters the agent's context, which matters because Cloudmersive keys access the entire OCR quota for the account.
Intent-based discovery
Agents search Jentic by intent (for example 'extract text from a PDF' or 'parse a business card') and Jentic returns the matching Cloudmersive operation with its input schema.
Time to first call
Direct Cloudmersive integration: half a day for the synchronous endpoints, longer if implementing async PDF job polling and multipart uploads. Through Jentic: under 1 hour — search, load schema, execute.
Related APIs
Alternatives and complements available in the Jentic catalogue.
FAQs
Specific to using Cloudmersive OCR API through Jentic.
What authentication does the Cloudmersive OCR API use?
The API uses an Apikey header. Generate the key in the Cloudmersive dashboard and pass it on every request as Apikey: <key>. Through Jentic, the key is stored encrypted in the vault and never enters the agent's context.
Can the Cloudmersive OCR API extract text from a PDF?
Yes. POST a PDF to /ocr/pdf/toText for plain text, or to /ocr/pdf/to/words-with-location and /ocr/pdf/to/lines-with-location for output with bounding box coordinates. Long documents use the async pattern via /ocr/pdf/get-job-status.
Does the Cloudmersive OCR API recognise business cards and forms?
Yes. /ocr/photo/recognize/business-card parses name, title, company, phone, and email from a card photo; /ocr/photo/recognize/form and /ocr/photo/recognize/form/advanced extract filled-in form values, with the advanced variant tuned for noisier captures.
How do I OCR an image through Jentic?
Run jentic search 'extract text from an image', load the Cloudmersive imageToText operation, and execute it with the source image. Jentic injects the Apikey header from the stored credential.
What are the rate limits for the Cloudmersive OCR API?
Rate limits depend on the Cloudmersive plan tier. The free tier allows a small monthly quota across all endpoints, while paid plans raise the per-minute and per-month ceilings. Check the Cloudmersive dashboard for the exact limits applied to your Apikey.
Can I get word-level coordinates for layout reconstruction?
Yes. /ocr/image/to/words-with-location and /ocr/pdf/to/words-with-location return each word with its bounding box, suitable for rendering highlight overlays or feeding NLP that needs spatial context.
GET STARTED