Image to text (OCR)
Read the printed text in a photo, scan or screenshot and get it back as plain text, in 12 languages. Finds the printed text in a picture and returns it as text and as a .txt file. Small pictures are enlarged and dark pictures are inverted first, because that makes the letters easier to read.
Input
Result
Choose a file and press Run.
Files you upload and the results are stored for one hour, then deleted. Anyone who has the link can open them, so don't upload anything private.
Use it from code or an agent
# 1. Upload the file. The reply ends with a line "ref <id>/f1": that is the file reference.
curl -T sample-scan.png https://setget.net/drop/
# 2. Run the tool with that reference. "wait" holds the reply for up to 30 seconds while the job runs.
curl -s https://setget.net/v1/tools/ocr-image \
-H 'Content-Type: application/json' \
-d '{"input":{"image":"DROP_ID/f1"},"wait":30}'
# 3. The reply is {"tool", "ms", "result", "files": [{"name", "size", "url", "ref"}], "job"}.
# Download a result file from its "url", or pass its "ref" to another tool.
# Still running after the wait? The reply is 202 with "job": follow it with
# curl -s https://setget.net/v1/jobs/JOB_ID?wait=30
import json, urllib.parse, urllib.request
BASE = "https://setget.net"
def call(method, path, body=None, headers=None):
request = urllib.request.Request(BASE + path, body, headers or {}, method=method)
return json.load(urllib.request.urlopen(request))
def upload(path):
"""Step 1: upload a file and get its reference."""
name = path.replace("\\", "/").rsplit("/", 1)[-1]
with open(path, "rb") as handle:
reply = call("PUT", "/drop/" + urllib.parse.quote(name), handle.read(), {"Accept": "application/json"})
return reply["file"]["ref"]
# Step 2: run the tool with the reference.
body = {"input": {'image': upload("sample-scan.png")}, "wait": 30}
reply = call("POST", "/v1/tools/ocr-image", json.dumps(body).encode(), {"Content-Type": "application/json"})
# Step 3: a job that is still running comes back as {"job": ...}. Wait for it.
while "result" not in reply and reply.get("job", reply).get("status") in ("queued", "running"):
reply = call("GET", "/v1/jobs/" + reply.get("job", reply)["id"] + "?wait=30")
if "error" in reply:
raise SystemExit(reply["error"]["message"])
print(reply["result"])
for file in reply["files"]: # result files: download from "url", or pass "ref" to another tool
print(file["name"], file["size"], file["url"], file["ref"])
const BASE = "https://setget.net";
// Step 1: upload a file and get its reference. file: a File or Blob
// (in Node: new Blob([fs.readFileSync(path)])).
async function upload(file, name) {
const response = await fetch(`${BASE}/drop/${encodeURIComponent(name)}`, {
method: "PUT", headers: {Accept: "application/json"}, body: file,
});
return (await response.json()).file.ref;
}
// Step 2: run the tool with the reference.
let reply = await (await fetch(`${BASE}/v1/tools/ocr-image`, {
method: "POST",
headers: {"Content-Type": "application/json"},
body: JSON.stringify({input: {"image":await upload(file, "sample-scan.png")}, wait: 30}),
})).json();
// Step 3: a job that is still running comes back as {job}. Wait for it.
while (!reply.result && ["queued", "running"].includes((reply.job || reply).status)) {
reply = await (await fetch(`${BASE}/v1/jobs/${(reply.job || reply).id}?wait=30`)).json();
}
if (reply.error) throw new Error(reply.error.message);
console.log(reply.result);
for (const file of reply.files) console.log(file.name, file.size, file.url, file.ref);
1. Upload. Tool: drop
Arguments:
{
"action": "create",
"files": [
{
"name": "sample-scan.png",
"size": "<bytes>"
}
]
}
PUT each file's bytes to its "upload.url" with the listed headers, and keep its "ref".
With a shell, one command does both: curl -T FILE https://setget.net/drop/
2. Run. Tool: run_tool
Arguments:
{
"slug": "ocr-image",
"input": {
"image": "DROP_ID/f1"
}
}
3. The answer holds "result" and "files" (each with "url" and "ref").
If it holds a "job" that is still running instead, follow it. Tool: job
Arguments:
{
"action": "wait",
"id": "JOB_ID"
}
{
"input": {
"image": "DROP_ID/f1"
},
"wait": 30
}
This is a file tool, so a call takes three steps: upload the file to get a reference, run the tool with that reference, then read the result and download its files. Uploads and results are public by link and expire after one hour.
Endpoint: POST /v1/tools/ocr-image. Over MCP: run_tool with slug: "ocr-image".
Machine-readable: input schema, Markdown. Connect an agent.
Inputs
| Name | Type | Required | What it does |
|---|---|---|---|
image |
file | yes | A picture that shows printed or on-screen text: PNG, JPEG, WebP, TIFF, BMP or GIF. In the API: a file reference such as DROP_ID/f1. Accepts: image. |
language |
string | no | Language of the text: eng (English), deu (German), fra (French), spa (Spanish), ita (Italian), por (Portuguese), nld (Dutch), nor (Norwegian), swe (Swedish), dan (Danish), pol (Polish), rus (Russian). One of: eng, deu, fra, spa, ita, por, nld, nor, swe, dan, pol, rus. Default: "eng". |
format |
string | no | text keeps the lines as they are in the picture. paragraphs puts each paragraph on one line. words also lists every word with its position and confidence. One of: text, paragraphs, words. Default: "text". |
layout |
string | no | How the text is arranged: auto works it out, single_column is one column of text, single_block is one even block, single_line is one line, sparse is scattered words such as labels on a photo or diagram. One of: auto, single_column, single_block, single_line, sparse. Default: "auto". |
invert |
string | no | Swap light and dark before reading. auto does it when the picture is mostly dark, such as a dark-mode screenshot. One of: auto, yes, no. Default: "auto". |
Examples
Read a scanned note
{
"image": "sample-scan.png"
}Press the example button above to run it on the sample file.
Get every word with its position
{
"image": "sample-scan.png",
"format": "words",
"layout": "single_block"
}Press the example button above to run it on the sample file.
Questions
- Which languages can it read?
- English, German, French, Spanish, Italian, Portuguese, Dutch, Norwegian, Swedish, Danish, Polish, Russian. Choose the language of the text: accented and special letters are only read correctly with the right one.
- How do I get the best result?
- Use a sharp, straight, evenly lit picture in which the letters are at least 20 pixels tall. Crop away everything that is not text. Screenshots and flat scans read best; photos taken at an angle read worst.
- Can it read handwriting?
- No. It reads printed and on-screen text. Handwriting usually comes out as nonsense.
- What does the confidence number mean?
- How sure the reader is, from 0 to 100, averaged over all words. Above 85 is usually almost error-free. Below 60, check the text carefully.
- What does the "words" format give me?
- A JSON file listing every word with its position in the picture in pixels (left, top, width, height), its line number and its confidence, as well as the plain text.
- Is my picture kept?
- No. It is read on our server and removed with the result when the download link expires.
Related tools
- Scanned PDF to text (OCR)Read the text in a scanned PDF and get it as a text file, or as a PDF you can search and copy from.
ocr-pdf - PDF to textGet the text out of a PDF, from all pages or the ones you choose, as plain text in reading order and as a .txt file.
pdf-text - Crop imageCrop an image to a rectangle in pixels, or to an aspect ratio such as 16:9 or 1:1 around its centre or an edge.
image-crop - Rotate and flip imageRotate an image by any angle and flip it horizontally or vertically. Quarter turns keep every pixel exact.
image-rotate-flip - QR code readerRead the text or link stored in a QR code from a picture or screenshot, and show where each code is.
qr-decode - Resize imageResize an image to a width, a height or both, keeping its proportions, cropping to fill, or stretching.
image-resize