A free, public REST API to convert and optimize raster + vector images. Single-file upload, fast turnaround, returns a download URL.
# Convert SVG to WebP
curl -X POST https://jinero.online/api/v1/images/convert \
-F "[email protected]" \
-F "format=webp" \
-F "preset=balanced"const form = new FormData();
form.append("image", file);
form.append("format", "webp");
form.append("preset", "balanced");
const r = await fetch("https://jinero.online/api/v1/images/convert", { method: "POST", body: form });
const { download_url, new_size } = await r.json();
console.log(download_url);All endpoints are public — no key required. Conversion and optimization are CPU-heavy, so they are rate-limited tighter than the Fonts API. Limits below are per IP.
# Anonymous limits (per IP):
# POST /convert 5 req/min
# POST /optimize 5 req/min
# GET /download/* 10 req/min
# GET /formats 60 req/min
# All responses include:
# X-RateLimit-Limit: 5
# X-RateLimit-Remaining: N
# Retry-After: seconds (on 429)Upload an image, get it back in a different format. Best for switching to modern formats (WebP, AVIF) for the web.
https://jinero.online/api/v1/images/convert
| Parameter | Type | Description |
|---|---|---|
image |
file (multipart) | Input file. Accepts jpg, jpeg, png, gif, bmp, webp, avif, heic, svg. Max 10 MB. |
format |
string | Output format: png, jpeg, webp, or avif. |
preset |
string | Quality preset: light, balanced (default), or strong. |
curl -X POST https://jinero.online/api/v1/images/convert \
-F "[email protected]" \
-F "format=webp" \
-F "preset=balanced"// Browser FormData example
const fileInput = document.getElementById("file");
const form = new FormData();
form.append("image", fileInput.files[0]);
form.append("format", "webp");
form.append("preset", "balanced");
const res = await fetch("https://jinero.online/api/v1/images/convert", {
method: "POST",
body: form,
});
const data = await res.json();
window.location = data.download_url;$response = Http::attach(
"image",
file_get_contents("photo.png"),
"photo.png"
)->post("https://jinero.online/api/v1/images/convert", [
"format" => "webp",
"preset" => "balanced",
]);
$result = $response->json();
$webpUrl = $result["download_url"];{
"success": true,
"session_id": "1d2e3f...",
"original_name": "photo.png",
"original_format": "png",
"output_format": "webp",
"original_size": 1024512,
"new_size": 156320,
"savings_percent": 84.7,
"dimensions": "1920x1080",
"preset": "balanced",
"download_url": "https://jinero.online/api/v1/images/download/1d2e3f...",
"expires_in_minutes": 60
}Compress without changing format. Useful for shrinking PNG/JPG files that are already in the right format but oversized.
https://jinero.online/api/v1/images/optimize
| Parameter | Type | Description |
|---|---|---|
image |
file (multipart) | Raster input (PNG, JPG, GIF, BMP, WebP, AVIF, HEIC). SVG is not accepted here — use /convert to rasterize first. |
preset |
string | light = near-lossless, balanced = recommended, strong = max compression. |
curl -X POST https://jinero.online/api/v1/images/optimize \
-F "[email protected]" \
-F "preset=strong"const form = new FormData();
form.append("image", file);
form.append("preset", "balanced");
const r = await fetch("https://jinero.online/api/v1/images/optimize", { method: "POST", body: form });
const { new_size, savings_percent } = await r.json();
console.log(`Saved ${savings_percent}%`);$res = Http::attach("image", file_get_contents("big.jpg"), "big.jpg")
->post("https://jinero.online/api/v1/images/optimize", ["preset" => "strong"]);
$data = $res->json();{
"success": true,
"session_id": "...",
"original_format": "jpeg",
"output_format": "jpeg",
"original_size": 2354821,
"new_size": 612491,
"savings_percent": 74.0,
"download_url": "https://jinero.online/api/v1/images/download/..."
}Convert raster images to clean SVG via VTracer. Best for logos, icons, and flat illustrations. Returns both raw SVG markup and a packaged ZIP.
https://jinero.online/api/v1/images/vectorize
| Parameter | Type | Description |
|---|---|---|
image |
file (multipart) | Raster input (jpg, jpeg, png, gif, bmp, webp). Max 10 MB. |
preset |
string | logo (default) · illustration · photo · sketch. Determines color quantization + path simplification. |
curl -X POST https://jinero.online/api/v1/images/vectorize \
-F "[email protected]" \
-F "preset=logo"
# Then fetch the SVG:
curl https://jinero.online/api/v1/images/vectorize/svg/SESSION_ID > out.svgconst form = new FormData();
form.append("image", file);
form.append("preset", "logo");
const r = await fetch("https://jinero.online/api/v1/images/vectorize", { method: "POST", body: form });
const { svg_url } = await r.json();
const svg = await fetch(svg_url).then(x => x.text());{
"success": true,
"session_id": "...",
"output_format": "svg",
"preset": "logo",
"dimensions": "1024x1024",
"svg_url": "https://jinero.online/api/v1/images/vectorize/svg/...",
"download_url": "https://jinero.online/api/v1/images/vectorize/download/..."
}AI-powered background removal (rembg / u2net family). Returns a transparent PNG. Max input 5 MB to keep latency reasonable.
https://jinero.online/api/v1/images/remove-background
| Parameter | Type | Description |
|---|---|---|
image |
file (multipart) | Raster input. Max 5 MB. |
preset |
string | standard (default, fast) · portrait (better hair edges) · quality · ultra (slowest, best). |
curl -X POST https://jinero.online/api/v1/images/remove-background \
-F "[email protected]" \
-F "preset=quality"const form = new FormData();
form.append("image", file);
form.append("preset", "portrait");
const r = await fetch("https://jinero.online/api/v1/images/remove-background", { method: "POST", body: form });
const { download_url } = await r.json();{
"success": true,
"session_id": "...",
"output_format": "png",
"preset": "quality",
"dimensions": "1920x1080",
"download_url": "https://jinero.online/api/v1/images/remove-background/download/..."
}AI 2× upscale via Real-ESRGAN. Only 2× is exposed publicly — higher scales (4×) and other modes (face restore, denoise, auto) remain UI-only for now due to GPU cost.
https://jinero.online/api/v1/images/upscale
| Parameter | Type | Description |
|---|---|---|
image |
file (multipart) | Raster input. Max 15 MB. Formats: jpg, jpeg, png, webp, bmp. |
scale |
integer | Only "2" is currently supported. Other modes (face, denoise, auto) are temporarily limited to the web UI. |
curl -X POST https://jinero.online/api/v1/images/upscale \
-F "[email protected]" \
-F "scale=2"const form = new FormData();
form.append("image", file);
form.append("scale", "2");
const r = await fetch("https://jinero.online/api/v1/images/upscale", { method: "POST", body: form });
const { download_url, dimensions } = await r.json();{
"success": true,
"session_id": "...",
"mode": "upscale",
"scale": 2,
"output_format": "png",
"dimensions": "3840x2160",
"download_url": "https://jinero.online/api/v1/images/upscale/download/..."
}Retrieve the binary result. Sessions stay valid for 60 minutes after creation; after that the file is purged. Each AI endpoint has its own download path (see Response examples).
https://jinero.online/api/v1/images/download/{session_id}
curl -L -o output.webp https://jinero.online/api/v1/images/download/1d2e3f...// Trigger browser download
window.location.href = data.download_url;Discover available output formats, presets, and limits at runtime — useful for building dynamic UIs that pick up new formats automatically.
https://jinero.online/api/v1/images/formats
curl https://jinero.online/api/v1/images/formatsconst r = await fetch("https://jinero.online/api/v1/images/formats");
const { output_formats, presets } = await r.json();{
"output_formats": [
{"key": "png", "name": "PNG", "extension": ".png", "supports_alpha": true},
{"key": "jpeg", "name": "JPEG", "extension": ".jpg", "supports_alpha": false},
{"key": "webp", "name": "WebP", "extension": ".webp", "supports_alpha": true},
{"key": "avif", "name": "AVIF", "extension": ".avif", "supports_alpha": true}
],
"presets": [
{"key": "light", "name": "Light", "description": "Near lossless, minimal compression"},
{"key": "balanced", "name": "Balanced", "description": "Best quality/size ratio"},
{"key": "strong", "name": "Strong", "description": "Maximum compression, smaller files"}
],
"supported_inputs": ["jpg","jpeg","png","gif","bmp","webp","avif","heic","svg"],
"max_file_size_bytes": 10485760,
"max_file_size_mb": 10,
"session_ttl_minutes": 60
}Standard HTTP semantics. Errors return JSON with a "message" field.
// 422 — validation (invalid format / unsupported file)
{ "message": "The image field must be a file of type: jpg, jpeg, png..." }
// 404 — download session not found or expired
{ "message": "Conversion result not found or expired" }
// 429 — rate limit exceeded
{ "message": "Too Many Attempts." }
// 500 — conversion failed (rare; details in error)
{ "message": "Failed to rasterize SVG. Make sure the file is valid." }About the jinero.online Images API
Public REST API for image conversion and optimization — convert SVG, JPG, PNG, HEIC into modern WebP or AVIF, compress without losing quality, free.
Modern formats
Output PNG, JPG, WebP, or AVIF. Convert legacy PNG/JPG hero images to WebP and watch LCP scores improve.
SVG → raster
Server-side SVG rasterization via rsvg-convert handles vector inputs cleanly — even SVGs with heavy filters.
No key required
Public endpoint, 5 req/min per IP for conversions. Generate an API token for higher limits when needed.
Simple JSON contract
Upload via multipart, receive a download URL valid for 60 minutes. No signing, no callbacks, no SDKs needed.
Frequently Asked Questions
No. The API is public and rate-limited per IP. /convert and /optimize are limited to 5 requests per minute; download and metadata endpoints have higher caps.
JPG, JPEG, PNG, GIF, BMP, WebP, AVIF, HEIC, and SVG. SVGs are rasterized server-side via rsvg-convert before encoding. Max file size: 10 MB.
PNG, JPEG, WebP, and AVIF. WebP gives the best size/quality ratio for the web; AVIF is even smaller but takes longer to decode. PNG keeps transparency; JPEG does not.
Originals are processed and immediately discarded. Converted results are kept in a temporary store for 60 minutes (long enough to follow the download_url), then deleted. We do not train any model on your uploads.
Those are specialized for one task (background removal or PNG compression). Our API focuses on format conversion + lossy/lossless compression with a single consistent interface, and is fully free without per-image fees or paywalls.