OCR.chat API

Jedan HTTP zahtjev pretvara sliku ili PDF u običan tekst, Markdown, tablice i JSON — na više od 100 jezika. Mjereno po stranici, bez iznenađenja.

Pregled

The OCR.chat API is a small REST interface. You POST a file and get back a job with the recognized text and a per-page breakdown (text, bounding boxes, confidence). Jobs of 5 pages or fewer return inline; larger jobs return immediately with a pending status that you poll until done.

  • Base URL: https://ocr.chat
  • Formats in: PNG, JPG, WEBP, GIF, BMP, TIFF, and multi-page PDF
  • Formats out: txt, md, docx, pdf, csv, json
  • Engines: cpu (fast, printed docs) and vlm (premium AI, handwriting, complex layout, math)

Autorizacija

Authenticate with your API token (find it on your account page) as a Bearer header:

Authorization: Bearer YOUR_API_TOKEN

You can also pass ?api_token=… as a query parameter. Usage is metered against your account's page balance.

Pošalji dokument

POST /api/v1/ocr/, multipart form upload.

curl -X POST https://ocr.chat/api/v1/ocr/ \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -F "file=@invoice.pdf" \
  -F "tier=vlm" \
  -F "language=auto"

Returns the job. For ≤5-page files it is already done with the text; larger files come back pending/processing, poll the status endpoint.

{
  "uuid": "9f2c1b7e4a...",
  "status": "done",
  "tier": "vlm",
  "language": "auto",
  "page_count": 1,
  "mean_confidence": 0.98,
  "text": "INVOICE\nAcme Corp\nTotal: 215.00 USD",
  "markdown": "# INVOICE\n\n**Acme Corp** ...",
  "pages": [ { "index": 0, "text": "...", "blocks": [ { "text": "...", "bbox": [x0,y0,x1,y1], "confidence": 0.98 } ] } ]
}

Dobiti rezultat

GET /api/v1/ocr/<uuid>/, poll until status is done or failed.

curl https://ocr.chat/api/v1/ocr/9f2c1b7e4a.../ \
  -H "Authorization: Bearer YOUR_API_TOKEN"

Preuzmi format

GET /api/v1/ocr/<uuid>/download/?format=md, export the result. format is one of txt, md, docx, pdf, csv, json.

curl -L "https://ocr.chat/api/v1/ocr/9f2c1b7e4a.../download/?format=docx" \
  -H "Authorization: Bearer YOUR_API_TOKEN" -o result.docx

Chat s dokumentom

Postavite pitanja o završenom zadatku. Odgovori se temelje samo na izvučenom tekstu i navode izvornu stranicu. Potrebno je imati račun - značajka razgovora je ograničena na račun.

POST /api/v1/chat/<uuid>/, JSON body {"message": "your question"}.

curl -X POST https://ocr.chat/api/v1/chat/9f2c1b7e4a.../ \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"message": "What is the invoice total and due date?"}'

Vraća pomoćnu poruku sa odgovorom i popisom citiranih stranica:

{"conversation": "a1b2…", "message": {
   "role": "assistant",
   "content": "The total is $42, due on March 3 (p. 1).",
   "citations": [{"page": 1, "snippet": "The invoice total is $42…"}]
}}

GET /api/v1/chat/<uuid>/history/, Dobiti potpuni transkript razgovora za posao.

Primjeri koda

import requests, time

API = "https://ocr.chat/api/v1/ocr/"
H = {"Authorization": "Bearer YOUR_API_TOKEN"}

# Submit
with open("invoice.pdf", "rb") as f:
    job = requests.post(API, headers=H,
        files={"file": f}, data={"tier": "vlm"}).json()

# Poll until done
while job["status"] in ("pending", "processing"):
    time.sleep(2)
    job = requests.get(API + job["uuid"] + "/", headers=H).json()

print(job["markdown"])

# Download as DOCX
r = requests.get(API + job["uuid"] + "/download/",
                 headers=H, params={"format": "docx"})
open("result.docx", "wb").write(r.content)
import fs from "fs";

const API = "https://ocr.chat/api/v1/ocr/";
const H = { Authorization: "Bearer YOUR_API_TOKEN" };

const form = new FormData();
form.append("file", new Blob([fs.readFileSync("invoice.pdf")]), "invoice.pdf");
form.append("tier", "vlm");

let job = await (await fetch(API, { method: "POST", headers: H, body: form })).json();

while (["pending", "processing"].includes(job.status)) {
  await new Promise(r => setTimeout(r, 2000));
  job = await (await fetch(API + job.uuid + "/", { headers: H })).json();
}
console.log(job.markdown);
# 1. Submit
curl -X POST https://ocr.chat/api/v1/ocr/ \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -F "file=@invoice.pdf" -F "tier=vlm"

# 2. Poll  (use the uuid from step 1)
curl https://ocr.chat/api/v1/ocr/UUID/ \
  -H "Authorization: Bearer YOUR_API_TOKEN"

# 3. Download
curl -L "https://ocr.chat/api/v1/ocr/UUID/download/?format=md" \
  -H "Authorization: Bearer YOUR_API_TOKEN" -o result.md

Parametri

FieldTypeDescription
filefileRequired. The image or PDF to process.
tierstringcpu (default, fast/printed) or vlm (premium AI: handwriting, layout, math).
languagestringauto (default) or a language code (en, ch, ja, ar, …).
toolstringOptional tool slug (e.g. extract-tables, handwriting-to-text) to apply that tool's preset.
translate_tostringFor the translate tool, target language code.

Greške i ograničenja

CodeMeaning
400No file, unsupported type, or file too large.
401Missing or invalid API token.
402Out of pages, daily/monthly free limit reached, or no credits. The body includes used/cap.
404Job UUID not found.
409Download requested before the job finished.

Each page processed costs credits (1/page on the fast tier, more on premium). Paid plans raise per-file page caps and add priority. See pricing.

Često postavljana pitanja

Create a free account and open your account page, your token is shown there with a copy button.

Yes, files of 5 pages or fewer return the full result inline in the POST response, so no polling is needed for most images and short PDFs.

Over 100, including Latin, CJK, Arabic, Cyrillic and Indic scripts. Use language=auto to detect, or pass a specific code.

Uploads are processed for OCR and deleted automatically. We never sell, share, or train on your documents.

Korištenje se mjeri po stranici protiv bilansa vašeg računa: anonimni pozivi dobivaju po IP dnevno dopuštenje, besplatni računi mjesečno dopuštenje, a plaćeni planovi koriste kupljene kredite sa većim ograničenjima i prioritetom po stranici po datoteci. Kada vam ponestane dobijete 402 sa iskorištenim i ograničenim u tijelu.

Možete poslati PNG, JPG, WEBP, GIF, BMP, TIFF i višestranični PDF. Rezultati se preuzimaju kao txt, md, docx, pdf (pretraživani), csv ili json putem formata parametra krajnje tačke preuzimanja.

400 je datoteka koja nedostaje, nepodržana vrsta ili datoteka prevelika; 401 nedostaje ili je nevažeća; 402 nema stranica; 404 nepoznat UUID posla; i 409 preuzimanje je zatraženo prije završetka posla. Tijela grešaka uključuju kratku poruku.

Objekt posla sa statusom, nivoom, jezikom, brojem stranica i srednjom pouzdanošću, plus puni tekst i oznake. Array stranica razbija svaku stranicu u blokove sa njihovim tekstom, okvirom za ograničenje (bbox), i pouzdanošću po bloku.

Koristi CPU (uobičajeno) za brzo, jeftino prepoznavanje čistih štampanih dokumenata. Koristi vlm, premium AI motor, za rukopis, kompleksne ili višekolonaške rasporede, matematiku i prijevod, gdje je mnogo precizniji.

Proslijedi alat sa oznakom (npr. extract-tables ili handwriting-to-text) da bi se primijenila podešena unaprijed postavljena postavka tog alata. Za alat za prevođenje, također proslijedi translate_to sa kodom ciljanog jezika da bi se dobio prepoznati tekst natrag preveden.

Datoteke od 5 stranica ili manje vraćaju se u POST odgovoru. Veće datoteke se vraćaju odmah kao čekanje ili obrada, a vi pitate GET /api/v1/ocr/<uuid>/ dok status nije završen ili neuspješan. Plaćeni planovi povećavaju ograničenje po datoteci.

API je jednostavan REST preko HTTPS-a, tako da radi iz bilo kojeg jezika sa HTTP klijentom - pogledajte Python, Node.js, i cURL primjere gore. Nema SDK-a za instaliranje; nekoliko linija standardnog HTTP koda je sve što vam treba.