anaksor-ocr
Sokheng · Kit
Reads Khmer and English text from documents, receipts and photos of signs, using a detector for text lines and a recogniser for each line. Suits developers who…
New
License plate number read from an uploaded photo
Newversion 1 · updated
Overview
An n8n workflow with a web form that takes a photo and returns the license plate number of the front-most car. A vision model on OpenRouter reads the plate.
License plate number read from an uploaded photo is an n8n workflow in the Productivity category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page or copy its JSON, then import it into an n8n canvas. Nothing runs until you activate it in n8n.
At a glance
- Runs in
- n8n Cloud: import the file into a new workflow there
- Your own n8n server: the same file and steps
- Another model: change the model name in Settings to any OpenRouter model that accepts images
- Another question: change the prompt in Settings, for example to read the text on a label
Platform
How it works
Trigger
A web form
It starts when someone submits the form:
- One JPG or PNG photo
- A required file field called Image
Actions
- The form’s photo is kept next to a model name and a prompt
- A chain sends the photo and the prompt to a vision model through OpenRouter
- The model is asked for only the characters of the plate on the front-most car
- The characters are shown on the form’s result page
Result
A result page that shows the plate characters
Integrations
What you need before you start.
OpenRouter
Vision model
Runs the model that reads the plate.
NeedsAn OpenRouter API key with credit, in an OpenRouter credential
Workflow check
Checked by Mindket on 13 Nov 2026
- Steps
- 4 steps, starting on a form
- AI
- 1 step calls an AI model: Basic LLM Chain
- Sign-ins
- Signs in to: OpenRouter
- API keys
- No API key is left in the file
- Step types
- Mindket’s checker reads all 4 of its step types
About
What it is
This n8n workflow reads a license plate from a photo. A web form takes one JPG or PNG. A Set node holds the model name and the prompt. A Basic LLM Chain sends the photo and the prompt to a vision model through OpenRouter. The model returns only the characters of the plate on the front-most car, and a result page shows them. Five nodes do the work.
Who it suits
Developers and small teams who want to see how a vision model reads text in a photo, such as a parking operator testing a plate reader. Skip it for anything that must be right every time. A model can misread a plate, so a person should check the result.
| Industry | Who uses it | How they use it | Example |
|---|---|---|---|
| Parking | Car park operators | Log the plates of cars photographed at a gate | An operator uploads a gate photo and notes the plate it returns. |
| Logistics | Yard managers | Check trucks arriving at a depot | A manager photographs a truck and compares the answer with the booking list. |
| Insurance | Claims staff | Read the plate from an accident photo | A handler uploads a claim photo and copies the plate into the case. |
| Development | Automation builders | Try a vision model on a real task | A developer swaps the model name in Settings to compare two models on the same photo. |
You need basic n8n skills and an OpenRouter account with some credit.
Where to use it
Any n8n that can reach OpenRouter will run it. The form page needs an address that people can open.
- n8n Cloud: import the file into a new workflow there
- Your own n8n server: the same file and steps
- Another model: change the model name in
Settingsto any OpenRouter model that accepts images - Another question: change the prompt in
Settings, for example to read the text on a label
Setup
Download saves a zip with workflow.json and Mindket’s setup notes in your downloads folder. One account is involved.
| Need | Detail |
|---|---|
| n8n | On n8n Cloud or on a server of your own |
| OpenRouter | An API key and some credit, saved in an OpenRouter credential |
| A photo | A JPG or PNG with a visible plate |
- In n8n, create a workflow, open its menu, choose Import from File and pick
workflow.json. - Add an OpenRouter credential and select it in
OpenRouter LLM. - Check the model name in
Settings. The file usesopenai/gpt-6-luna. - Open
FromTrigger, copy its test address and upload a photo. - Activate the workflow to get the production form address, which
FromTriggershows.
How to use it
Open the form address, choose a JPG or PNG and submit it. The page that comes back shows the plate characters.
| To change | Where |
|---|---|
| The model | The model value in Settings |
| What it is asked to read | The prompt value in Settings |
| The accepted file types | acceptFileTypes in FromTrigger |
| The result page title | completionTitle in FormResultPage |
Example
A parking operator uploads a photo of a car at a gate.
The model is told to return only the characters, so the page should show a short string such as the plate number. If no plate is readable, the answer can be empty or wrong.
For newcomers and experts
If you are new. n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. A vision model is an AI that can look at a picture and answer questions about it. OpenRouter is one service that gives access to many such models.
For experts. The photo reaches the chain as binary data in the field Image, which the form names after its field label. The Set node keeps binary data, so the chain’s image message can read it. The model name comes from $json.model in OpenRouter LLM.
| Node | What it does | What to know |
|---|---|---|
FromTrigger | A form with one required file field called Image | The uploaded photo lands in the binary field Image |
Settings | Adds the model and prompt fields | It keeps the other fields, including the photo |
Basic LLM Chain | Sends the prompt and the photo to the model | Its image message reads the binary field Image |
OpenRouter LLM | Provides the model chosen in Settings | The model is read from the field model of each item |
FormResultPage | Shows the answer on the form’s result page | It shows the chain’s text field |
Good to know
The model was replaced. Mindket did not run this workflow.
- Model: the file used
openai/gpt-4o. OpenAI’s deprecations page, read on 12 November 2026, shuts the datedgpt-4osnapshots down on 14 January 2027, withgpt-6-lunaas the replacement. Mindket’s copy usesopenai/gpt-6-luna, which OpenRouter’s public list, read on the same day, shows with image input at 0.10 US dollars per million input tokens and 0.50 per million output tokens - Plain chain:
gpt-6-lunacalls tools only in the Responses API, or with reasoning effortnoneon Chat Completions (OpenAI’s model page), and this chain uses no tools - Privacy: a license plate can identify a person in many places. Check the rules that apply to you before you process photos of other people’s cars, and keep the form address private
- Accuracy: a model can misread a plate or a photo with glare. A person should check the result
- Your data: the photo goes to OpenRouter and the model’s provider. The file holds no credential
README and files
License plate number read from an uploaded photo
An n8n workflow with a web form that takes a photo and returns the license plate number of the front-most car. A vision model on OpenRouter reads the plate.
What is inside
workflow.json: the workflow as n8n exports it, five nodes. It holds no credential and no sticky notes.README.md: these notes.
What we changed from the original copy
- The model was
openai/gpt-4o. OpenAI’s deprecations page, read on 12 November 2026, shuts the datedgpt-4osnapshots down on 14 January 2027, withgpt-6-lunaas the replacement. OpenRouter’s public list still showsopenai/gpt-4owithout an end date. Mindket’s copy usesopenai/gpt-6-luna, which the list shows with image input. - The form trigger and the result page carried webhook IDs from the author’s n8n. They are gone, so n8n makes new ones when you import the file.
- The OpenRouter node was linked to a credential in the author’s n8n. The link is gone, so select your own.
- Nothing else. The prompt, the form field and the result page are the file’s own.
Before you start
- An n8n instance, on n8n Cloud or your own server.
- An OpenRouter API key with some credit.
- A JPG or PNG photo with a visible license plate.
Set it up
- In n8n, create a workflow, open its menu, choose Import from File and pick
workflow.json. - Add an OpenRouter credential and select it in
OpenRouter LLM. - Check the model name in
Settings. The file usesopenai/gpt-6-luna. - Open
FromTrigger, copy its test address and upload a photo. - Activate the workflow to get the production form address, which
FromTriggershows.
How a run goes
FromTriggershows a form with one required file field called Image, and starts the run when someone submits it.Settingsadds the model name and the prompt to the item and keeps the photo.Basic LLM Chainsends the prompt and the photo to the model thatOpenRouter LLMprovides.FormResultPageshows the model’s answer on the page that comes back.
Change the result
- The model: The
modelvalue inSettings. - What it is asked to read: The
promptvalue inSettings. - The accepted file types:
acceptFileTypesinFromTrigger. - The result page title:
completionTitleinFormResultPage.
Good to know
- Mindket did not run this workflow. OpenRouter was not called, and nobody tried a real photo. The settings come from n8n’s source and OpenRouter’s and OpenAI’s pages, read on 12 November 2026.
- The form names the uploaded file after its field label, so the photo sits in the binary field
Image. The chain’s image message reads that field. - OpenAI’s model page says
gpt-6-lunacalls tools in the Responses API, or on Chat Completions only with reasoning effortnone. This chain uses no tools. - OpenRouter’s public list shows
openai/gpt-6-lunaat 0.10 US dollars per million input tokens and 0.50 per million output tokens. - A license plate can identify a person in many places. Check the rules that apply to you before you process photos of other people’s cars.
- A model can misread a plate. A person should check the result.
- The photo goes to OpenRouter and the model’s provider.
- The file holds no credential and no key.
NOTICE.md 643 B README.md 3 KB workflow.json 4 KB
What’s new
Reviews
More from Mindket
See all from MindketQwen2.5-7B-Instruct-GGUF by Qwen
Qwen · Kit
Qwen2.5-7B-Instruct as GGUF files in eight quantization sizes, to chat with locally through llama.cpp. For people who want a capable 7.6B model on their own co…
New
ACE-Step 1.5 XL SFT
ACE-Step · Kit
Generates music from text with a 4B diffusion decoder, 50 steps and guidance control, aiming for higher audio quality than the base. For creators with a GPU of…
New
Similar in Productivity
See all in ProductivityGuizang PPT Skill
op7418 · Skill
Skill that makes single-file HTML slide decks, slide illustrations and platform covers in two visual styles, with a presenter mode for rehearsal. For speakers…
New29.8k
Frontend Slides
zarazhangrui · Skill
Builds HTML presentations from scratch or from a PowerPoint file and shows visual style previews to pick from, for people who want web slides with no CSS or Ja…
New32.3k
IPTV
iptv-org · Kit
Collects publicly available live TV channels from many countries into playlists that any video player with live streaming can open. Made for viewers and for de…
New144k