Classify images
Give your own list, such as kitchen, bathroom, exterior and living room, and each photo comes back in one of those categories with how confident the match is. Sure calls can go straight through and doubtful ones to a person. You pay per use, with no subscription.
Run it online
Run this on our servers with your account. Free tools run in your browser; this one bills your KIT balance per the price above.
sort them into your own categories
Knowing which bucket a photo belongs to
Knowing what is in a photo and knowing where it belongs are two different questions. A marketplace does not need to hear that an image shows wood, grain, brown and furniture; it needs to know the listing goes under Furniture, then Tables, so it reaches the right review queue or the right section of the store. Classification answers that second question: one image goes in, one label comes out, picked from the categories you defined for your catalog, your workflow or your moderation rules. Each answer also says how confident the match is, which is what lets you decide when a person should look again.
How teams put classification to work
The usual pattern is simple: classify each photo as it arrives, send it on by category, and pass the low-confidence results to someone for manual review. A real estate site can sort property photos into kitchen, bathroom, exterior and living room to build tidy galleries. An insurer can split damage photos into vehicle, property or document so each claim lands in the right queue, and a creative team can file incoming assets under its own campaign names. When you also want descriptive keywords, pair it with a tagging tool, because a category is a decision, not a list of loosely related words.
Sorting, not describing
There's a real difference between knowing what's in a photo and knowing which bucket it belongs to. A marketplace doesn't need to know an image contains 'wood, grain, brown, furniture' — it needs to know the listing belongs in 'Furniture > Tables', full stop, so it routes to the right review queue or the right storefront section. image.classify is built for that decision: one image in, one category label out, chosen from the set you define for your own catalog, workflow or moderation rules.
Fitting it into automation
Classification pairs naturally with a moderation or listing pipeline: classify on upload, auto-route by category, and escalate low-confidence results for manual review. Combine it with image.tag when you also want descriptive keywords alongside the category, or with image.quality upstream so garbled uploads never reach the classifier at all. Pricing is a small per-request base plus a per-image rate, both published; a failed request is retried up to three times and, if it still fails, costs nothing.
What you can do with it
Marketplace listing routing
Classify seller photos into your category tree automatically so listings land in the right section without manual review.
Real estate photo sorting
Sort property photos into kitchen, bathroom, exterior and living room automatically to build organized listing galleries.
Insurance claim triage
Classify submitted damage photos into vehicle, property or document types to route each claim to the right queue.
Content pipeline routing
Sort incoming creative assets into your brand's own campaign categories before they reach a design or publishing team.
FAQ
Is image classification free to use?
There's no free tier — free tiers get abused and slow everyone down. No subscription, no tokens, no invented credits, and a failed task is never charged.
How confident is the classification result?
Each result includes a confidence score for the chosen category, so you can set your own threshold for auto-approval versus human review.
What's the difference between classification and tagging?
Classification assigns one image to one category from your list; image.tag instead returns multiple descriptive keywords, ranked, with no fixed category set.
What happens if a classification request fails?
It's retried automatically up to three times; if it keeps failing you receive a clear error and are never billed for it.
Can I use my own categories instead of a fixed list?
Yes. You write the list for each job, as broad as indoor and outdoor or as specific as your own catalog tree, and every image is placed in the one that matches best.
Related tools
For developers — API access
Everything on this page is available programmatically. This section is for teams who want to wire it into their own systems; everyone else can just use the tool above.
API endpoint
Prefer to automate it? One authenticated POST creates the task; the result comes back by webhook or a signed link. The same capability also runs here on the web, by email and from Telegram — and soon from our app too.
Call it from your stack
curl -X POST https://api.kit.forhosting.com/image/classify \
-H "Authorization: Bearer $KIT_KEY" \
-H "Content-Type: application/json" \
-d '{"image":"https://ejemplo.com/imagen.jpg"}'const res = await fetch("https://api.kit.forhosting.com/image/classify", {
method: "POST",
headers: {
"Authorization": `Bearer ${process.env.KIT_KEY}`,
"Content-Type": "application/json"
},
body: JSON.stringify({
"image": "https://ejemplo.com/imagen.jpg"
})
});
const { task_id } = await res.json();import os, requests
res = requests.post(
"https://api.kit.forhosting.com/image/classify",
headers={"Authorization": f"Bearer {os.environ['KIT_KEY']}"},
json={
"image": "https://ejemplo.com/imagen.jpg"
},
)
task_id = res.json()["task_id"]<?php
$res = file_get_contents("https://api.kit.forhosting.com/image/classify", false, stream_context_create([
"http" => [
"method" => "POST",
"header" => "Authorization: Bearer " . getenv("KIT_KEY") . "\r\nContent-Type: application/json",
"content" => '{"image":"https://ejemplo.com/imagen.jpg"}',
],
]));
$task = json_decode($res, true);body := bytes.NewBufferString(`{"image":"https://ejemplo.com/imagen.jpg"}`)
req, _ := http.NewRequest("POST", "https://api.kit.forhosting.com/image/classify", body)
req.Header.Set("Authorization", "Bearer "+os.Getenv("KIT_KEY"))
req.Header.Set("Content-Type", "application/json")
res, _ := http.DefaultClient.Do(req)Example request
{
"image": "https://ejemplo.com/imagen.jpg"
}Example response
{
"task_id": "tsk_a1b2c3d4e5f6a1b2c3d4e5f6",
"type": "image.classify",
"status": "queued",
"_links": {
"result": "/tasks/tsk_…/result"
}
}The API is asynchronous: the call returns a task_id immediately and the result arrives by webhook. Polling is capped at 1 req/s per task.
Pricing
Published price — no tokens, no invented credits. A failed task is never charged.
Limits
max_mb | 15 |
max_megapixels | 12 |
Errors
| HTTP | Code | Meaning |
|---|---|---|
401 | unauthorized | Missing or invalid API key. |
402 | insufficient_balance | Your balance doesn't cover the task price. |
404 | unknown_type | That task type doesn't exist. |
429 | rate_limited | Too many requests. Use the webhook instead of polling. |
422 | task_failed | The task failed after 3 retries. You are never charged for it. |
Technical notes
Define the categories that matter to your product — not a generic taxonomy someone else picked — and this endpoint sorts every incoming image into one of them. It's the routing decision behind automated triage, not a list of loosely related keywords.
How you shape the categories
The result comes back as the best-matching category along with how confident the match is, so downstream logic can auto-approve high-confidence calls and route the uncertain ones to a human. Call POST /image/classify with the image and your list of candidate categories — the labels can be as broad as 'indoor / outdoor' or as specific as your internal SKU taxonomy.
Async by design, at any volume
A single product upload is one call; a catalog migration is tens of thousands. Either way the endpoint answers immediately with a task_id and classifies in the background, so your application never blocks on a slow image model. Results arrive by webhook the moment they're ready, or you pull them from a signed link that stays valid for 24 hours.
A discipline with a long lineage
Sorting images into predefined buckets is one of the oldest problems in computer vision, going back to early digit and object-recognition research decades before anyone called it 'machine learning'. What's changed is accessibility: what once required a research team and custom-trained models is now a single POST request with your own category list, no model training pipeline of your own required.
Technical questions
Can I define my own categories for image classification?
Yes — you send your own list of candidate categories with each request; the API isn't limited to a fixed generic taxonomy.
How is image.classify priced?
A small fixed fee per request plus a per-image rate, both published on this page, with no invented tokens or credits.
Can I classify a large batch of images?
Yes. Each call is an independent async task, so you can submit large batches and collect results as each one completes.
How do results come back?
By signed webhook, which is recommended for pipelines, or through a signed result link that remains valid for 24 hours.