What does an image to line art converter do?

It finds the edges in your photograph and renders them as strokes, discarding colour, texture and shading. The result is a line drawing whose geometry matches the original, because it is extracted from it rather than drawn again.

Which method should I choose?

Line Art for a drawing you will look at, and it is the default. TEED when small detail matters, PIDI for people and products, HED when you want soft contours and some texture kept. M-LSD only for buildings and interiors, since it finds straight lines and ignores organic shapes. Scribble when you want a loose hint rather than a faithful trace.

Can I download the result as SVG or a vector file?

No. The output is a raster image — PNG-style pixels, not paths. If you need vector output for cutting, laser work or infinite scaling, run the result through a separate raster-to-vector tracer, or use a tool built for vectorising.

Why use this when there are free line art converters?

Often you should not, and it is worth being straight about that. Several free tools run entirely in your browser, cost nothing and never upload your photo, and for a high-contrast image a simple threshold does a perfectly good job. What you get here is six neural detectors that hold contours where a threshold breaks up — soft hair, low contrast, dense overlapping detail — plus the result landing in a session next to the models that will consume it.

How is this different from an AI line drawing generator?

A generator draws a new picture in a line style, guided by your photo. This traces the photo itself. If you want the proportions and placement of the original to survive exactly, you want this; if you want an illustration that merely resembles it, you want the generator.

Can I use the output as a ControlNet lineart input?

Yes, that is what most people convert for. All six methods produce standard edge-map images that work anywhere a lineart or scribble conditioning input is expected. Because the result opens in a generation session, you can wire it into a model here without downloading it first.

What image formats can I upload?

JPG, PNG and WebP, one image per run. The line art comes back as a standard image you can download or pass straight to another node.

Image to line art converter

Image to Line Art: Convert a Photo into a Line Drawing

Upload a picture and get its outlines back as line art, with the colour and shading stripped away. Six methods, from a clean drawable line to a loose scribble — and every one of them traces the photo you gave it rather than redrawing it from scratch.

Line extraction method

Clean drawable outlines. The default.

2 credits

Portrait photograph of a woman with thick curly hair against a plain wall

Line art of the same portrait from Line Art: fine continuous strokes tracing individual curls and the knitted collar

Examples

Line Art Examples

Line Art — the default
HED — soft contours
PIDI — smoothest
TEED — finest detail
M-LSD — straight lines only
Scribble — sparsest

How it works

Convert an Image to Line Art in Three Steps

Step 1
Upload a photo
Drop in one JPG, PNG or WebP. Good contrast between the subject and its surroundings gives the cleanest lines; a low-contrast or very dark photo gives the detector less to find.
Step 2
Pick a method
Six detectors, and they genuinely disagree about what counts as an edge. Line Art is the default and looks most like a drawing; the section below says when to reach for the others.
Step 3
Download or keep working
The line art opens in a generation session, so you can download it or wire it straight into a model as a lineart conditioning input without starting again.

Choose a method

Six Ways to Extract Line Art

Every method costs the same, so trying a second one on the same photo is cheap. They differ in how much detail survives, not in quality.

Line Art

Clean, drawable outlines for illustration, anime and design sketches. The default, and the most legible of the six at a glance.

HED

Soft holistic contours for architecture, interiors and products. Keeps texture and tonal boundaries that the clean-line methods throw away.

PIDI

The smoothest of the six. It holds the overall shape of a subject and lets fine detail go, which suits a soft-edged portrait or a product contour more than a busy scene.

TEED

Fine dense edges for complex characters, products and mechanical detail. The most literal of the six, and the one to use when small features matter.

M-LSD

Straight-line and perspective detection for rooms and architecture. It finds structural lines and ignores organic shapes, which makes it the wrong choice for a portrait and the right one for a building.

Scribble

Sparse sketch guidance for rapid layout and early concept design. Deliberately loose — it gives a generation model room to invent rather than pinning every contour.

What this is

It Traces Your Photo, It Does Not Redraw It

The geometry survives

Edges are extracted from the pixels you uploaded, so proportions, angles and placement come through unchanged. An AI illustration tool that draws a new picture in a line style cannot promise that, and this is the difference worth knowing before you pick one.

Detail is a choice, not a quality

Scribble finding fewer lines than TEED is not Scribble doing a worse job. The six sit on a scale from sparse to literal, and which end you want depends on whether the lines are the finished artwork or a hint for a model.

Colour and shading do not survive

Everything tonal is discarded by design — that is what makes the output usable as conditioning. The result is WHITE LINES ON BLACK, which is the standard form a conditioning input takes; invert it in any image editor if you want dark lines on paper. A photo that relies on colour rather than contrast to separate its subject will also give thinner lines than you expect.

What it is for

What People Use Line Art For

ControlNet lineart conditioning

The most common use. Line art hands a generation model the exact structure of a photo, so a new image keeps the composition while its style changes completely.

Colouring pages and inking

A clean outline pass is the starting point for a colouring sheet or a digital ink layer, with the shading left for you to put back.

Product and fashion flats

Stripping a product photo to its contours gives you the technical-flat view that a catalogue or a spec sheet wants.

Architectural line drawings

M-LSD finds the structural lines in a building photo and ignores the rest, which is closer to an elevation drawing than a filter.

Credits and plans

Each conversion costs a flat 2 credits, whatever the size of the photo or the method you pick.

Basic

For hobbyists and explorers

$9.99/mo

Select Plan

1000 credits
Up to ~200 images/month
At least ~1000s video/month
Intelligent creative agent
Video editor
Latest image models, including GPT Image 2.5, GPT Image 2 and Nano Banana Pro
Latest video models, including Seedance 2
Realistic face uploads
Voice generation with ElevenLabs
No watermark
Unlimited template access

Pro

Max

For power users and teams

$149.99/mo

Select Plan

18000 credits
Up to ~6000 images/month
At least ~18000s video/month
Intelligent creative agent
Video editor
Latest image models, including GPT Image 2.5, GPT Image 2 and Nano Banana Pro
Latest video models, including Seedance 2
Realistic face uploads
Voice generation with ElevenLabs
No watermark
Unlimited template access

Team

For teams and studios

$329.99/mo

Select Plan

36300 credits
Up to ~12000 images/month
At least ~36300s video/month
Realistic face uploads
Share canvas, workflows, and assets with your team
Up to 10 members per team
Up to 5 teams
Role-based management
Team credit management and spending caps
Per-member usage tracking

See all plans

FAQ

Image to Line Art Questions

What does an image to line art converter do?: It finds the edges in your photograph and renders them as strokes, discarding colour, texture and shading. The result is a line drawing whose geometry matches the original, because it is extracted from it rather than drawn again.
Which method should I choose?: Line Art for a drawing you will look at, and it is the default. TEED when small detail matters, PIDI for people and products, HED when you want soft contours and some texture kept. M-LSD only for buildings and interiors, since it finds straight lines and ignores organic shapes. Scribble when you want a loose hint rather than a faithful trace.
Can I download the result as SVG or a vector file?: No. The output is a raster image — PNG-style pixels, not paths. If you need vector output for cutting, laser work or infinite scaling, run the result through a separate raster-to-vector tracer, or use a tool built for vectorising.
Why use this when there are free line art converters?: Often you should not, and it is worth being straight about that. Several free tools run entirely in your browser, cost nothing and never upload your photo, and for a high-contrast image a simple threshold does a perfectly good job. What you get here is six neural detectors that hold contours where a threshold breaks up — soft hair, low contrast, dense overlapping detail — plus the result landing in a session next to the models that will consume it.
How is this different from an AI line drawing generator?: A generator draws a new picture in a line style, guided by your photo. This traces the photo itself. If you want the proportions and placement of the original to survive exactly, you want this; if you want an illustration that merely resembles it, you want the generator.
Can I use the output as a ControlNet lineart input?: Yes, that is what most people convert for. All six methods produce standard edge-map images that work anywhere a lineart or scribble conditioning input is expected. Because the result opens in a generation session, you can wire it into a model here without downloading it first.
What image formats can I upload?: JPG, PNG and WebP, one image per run. The line art comes back as a standard image you can download or pass straight to another node.