ocr-api-simple web app (for thai/en lang.)

using Typhoon OCR

A single-file browser tool to extract raw text from images and PDFs using Typhoon OCR.

No install. No server. Just open index.html.

---

Setup

1. Get a Typhoon API key → opentyphoon.ai
2. Open index.html in your browser
3. Paste your API key into the input field

Keys are never stored — session memory only.

---

Usage

1. Enter your API key
2. Upload an image (PNG, JPG, WEBP) or PDF
3. Click Extract Text
4. Copy the result

---

OCR Parameters

Tweak these inside extractTextFromImage() in index.html:

Parameter	Default	Description
model	typhoon-ocr	OCR model
task_type	default	Extraction mode
max_tokens	16384	Max output tokens
temperature	0.1	Lower = more deterministic
top_p	0.6	Nucleus sampling threshold
repetition_penalty	1.0	Penalises repeated tokens
pages	null	Specific pages to process (PDF), null = all

---

Rate Limits (typhoon-ocr)

Limit	Value
Requests/second	2
Requests/minute	20

Need higher limits? Email contact@opentyphoon.ai.