📸 OCR Label Scan

Photograph the label.
We'll read it.

Snap a photo of any appliance nameplate. Our GPT-4o Vision pipeline extracts brand, model, and serial number in one shot — then passes it directly to the lookup engine. No manual transcription, no typos, no wrong model.

SAMSUNG

MODEL

WF45T6000AW

SERIAL

0H9E4ABTA

✓

Brand extracted

Samsung · confidence HIGH

✓

Model extracted

WF45T6000AW

✓

Serial extracted

0H9E4ABTA00271L

→ Passed to full lookup pipeline

Features

Works on any label.
Any condition.

🤖

GPT-4o Vision

Not a basic OCR library — GPT-4o Vision understands appliance label layouts across every major manufacturer format. Handles rotation, perspective, and partial occlusion.

📱

Mobile-first

Designed for phone cameras in real-world conditions — closets, behind machines, dim laundry rooms. Works even when the label is partially faded or worn.

⚡

One-shot extraction

One image → brand, model, serial extracted together. Then passed automatically to the full lookup pipeline. You never manually type a model number again.

🔬

Confidence scoring

Each extracted field comes with a confidence score. Low-confidence extractions are flagged so you can verify before proceeding — not silently wrong.

🔗

API endpoint

POST an image to /v1/ocr/label — get brand, model, serial back. Chain to /v1/lookup for the full profile.

🌐

Multi-language labels

Handles labels in English, Spanish, French, German, Korean, and Chinese — covering every major manufacturing market's label format.

API integration

Two endpoints.
Full pipeline.

Send an image to OCR, get model + serial back. Pass those to lookup for all 12 data points. The entire pipeline in two API calls.

Available on Team, Business, and Enterprise plans. 100 OCR scans/mo on Team, 1,000 on Business.

# Step 1: Extract from image
response = requests.post(
  "https://api.applianceiq.com/v1/ocr/label",
  headers={"X-API-Key": key},
  files={"image": open("label.jpg", "rb")}
)
ocr = response.json()
# ocr.model_number → "WF45T6000AW"
# ocr.serial_number → "0H9E4ABTA00271L"
# ocr.confidence → "HIGH"
# Step 2: Full lookup
lookup = requests.post(
  "https://api.applianceiq.com/v1/lookup",
  headers={"X-API-Key": key},
  json={
    "model_number": ocr["model_number"],
    "serial_number": ocr["serial_number"]
  }
).json()
# lookup.recalls → []
# lookup.manufacturing_date → "2021/05"

Photograph the label.We'll read it.

Works on any label.Any condition.

Two endpoints.Full pipeline.

Never type a model numberagain.

Photograph the label.
We'll read it.

Works on any label.
Any condition.

Two endpoints.
Full pipeline.

Never type a model number
again.