Pencil sketch
A graphite drawing built from a color-dodge outline layer and shading strokes that follow the direction of edges in your photo, on subtle paper grain. Best for portraits and architecture.
Image tools · Free
Turn your photos into beautiful pencil sketches, cartoons, and artistic illustrations instantly. Completely free, no signup, no watermark, and processed directly in your browser.
Working…
The Photo to Sketch & Cartoon Converter is a free online tool from Entertainment Couch that turns a photo into a pencil sketch, cartoon, line drawing, comic, anime-style illustration, watercolor, oil painting or pop art in your web browser. It uses real image-processing algorithms and optional on-device AI models, needs no account, adds no watermark, and never uploads your photos.
| Price | Free. Unlimited photos, no watermark and no premium version. |
|---|---|
| Account | None. No sign-up, email or app download. |
| Privacy | Photos are processed in your browser and are never uploaded or stored on a server. |
| Styles | 14 algorithmic styles (pencil, color pencil, charcoal, ink, line art, cartoon, comic book, anime-inspired, watercolor, oil, pop art, digital illustration, portrait cartoon, pet cartoon) and 7 AI neural painting styles plus your own style image. |
| On-device AI | Subject detection (U²-Net), face detection (UltraFace) and neural style transfer (Magenta). Optional, downloaded once, run locally. |
| Controls | Intensity, detail level, line weight, subject and background styling, selection brush, before/after slider, zoom, pan, undo and redo. |
| Input | JPG, PNG, WebP, GIF (first frame), BMP and AVIF; HEIC where your browser can open it. Up to 50 MB per file. |
| Output | PNG (with transparency), JPG or WebP at original or reduced size. Batch ZIP for several photos. |
| Devices | Current Chrome, Edge, Firefox and Safari on Windows, macOS, Linux, Android, iPhone and iPad. |
| Made by | Entertainment Couch. Last updated November 8, 2026. |
Add your photo, pick a style, adjust the intensity and detail, check the before/after preview, and press Download. Everything runs in your browser, so there is no upload, account or software to install.
Drop it on the tool, press Choose photos, or paste it with Ctrl+V or ⌘V. Add several to process a batch.
Every style tile shows a live preview of your own photo, so you can compare them at a glance.
Set the intensity, detail level and line weight. Style the subject and background separately if you like.
Use the Compare slider, zoom in to check details, and undo or redo any change.
Save PNG, JPG or WebP at full resolution with no watermark, or save a whole batch as a ZIP file.
You can create pencil, color pencil and charcoal sketches, ink drawings, line art, cartoons, comic book art, anime-inspired illustrations, watercolors, oil paintings, pop art and polished digital illustrations, plus cartoon modes tuned for portraits and pets and seven AI neural painting styles.
A graphite drawing built from a color-dodge outline layer and shading strokes that follow the direction of edges in your photo, on subtle paper grain. Best for portraits and architecture.
Colored pencil strokes laid over white paper, with more pigment where the photo is darker or more saturated. Gives a hand-drawn, airy look.
Deep, smudged tones on warm toned paper. The charcoal skips the “valleys” of the paper texture, just like the real medium.
Crisp black outlines plus up to three layers of cross-hatching for shadows. Pure black and white, ready for printing or tattoo stencils.
Clean outlines only, extracted with flow-guided edge detection. Ideal for coloring pages, tracing and minimalist prints.
Smooth, flat color regions from edge-preserving filtering and color quantization, finished with bold outlines and brighter colors.
Fewer, punchier colors, thick inked outlines, solid black shadows and halftone dot shading in the midtones.
Soft cel shading with two tone steps, cool shadows, a gentle highlight bloom and clean tinted line work. It creates an original look and does not copy any character or artist.
Transparent washes with irregular borders, pigment pooling along edges, granulation and paper texture, over a faint pencil underdrawing.
Paint blocked in with a Kuwahara filter, then brush strokes and impasto relief that follow the forms, lit from the top left on canvas texture.
Bold posterized bands mapped to vivid palettes, halftone dots and heavy outlines, with an optional 2 × 2 grid of four color variations.
Rich color with soft gradients kept, local contrast, a light glow and refined colored line work, like a polished digital painting.
A cartoon tuned for people: faces are found on your device and keep finer lines and shading so the person stays recognizable, with a bold outline around the subject.
A cartoon tuned for animals, with fur strokes that follow the direction of the coat, warmer colors and a bold outline around your pet.
Seven painting styles generated by a neural style-transfer network, from swirling night-sky brushwork to stained-glass mosaic, plus the option to use your own style image.
Keep the background as a photo, blur it, turn it black and white, replace it with a solid color, remove it, or give it a different style from the subject.
For portraits, use Pencil Sketch, Portrait Cartoon or Anime-Inspired. For pets, use Pet Cartoon or Color Pencil. For landscapes and buildings, use Watercolor, Oil Painting or Ink Drawing. For coloring pages, use Line Art.
| Photo type | Best styles | Recommended settings |
|---|---|---|
| Portraits and selfies | Pencil Sketch, Portrait Cartoon, Anime-Inspired, Digital Illustration | Detail: Balanced or High; keep AI assist on so faces are preserved |
| Pets and animals | Pet Cartoon, Color Pencil, Charcoal, Watercolor | Detail: High for fur; Background: Soft blur or Solid color |
| Landscapes and travel | Watercolor, Oil Painting, AI neural painting | Detail: Balanced; Intensity 60 to 80 |
| Buildings and products | Ink Drawing, Line Art, Comic Book | Detail: High; Line weight 80 to 120% |
| Coloring pages and stencils | Line Art, Ink Drawing | Detail: Minimal for bold, simple outlines |
| Posters and social media | Pop Art, Comic Book, Cartoon | Try the 2 × 2 pop art grid and Solid color backgrounds |
The converter analyzes your photo’s edges, colors and textures with established computer-vision algorithms, then redraws it with simulated pencil, ink, paint or cartoon techniques. Optional AI models running on your device find the subject and faces and power the neural painting styles.
Sobel gradients and a smoothed structure tensor measure where edges are and which way they run. Lines come from a flow-guided difference of Gaussians, which gives smooth, connected strokes instead of noisy pixels [1].
A bilateral filter flattens textures while keeping outlines sharp [2], and a Kuwahara filter turns areas into paint-like patches for the oil style [3].
k-means clustering picks a small palette that represents your photo, which creates the flat color regions of cartoons and comics.
Line integral convolution draws pencil, charcoal, fur and brush strokes along the direction of the image [4], combined with paper grain, halftone screens and lighting for impasto.
U²-Net finds the main subject [5] and UltraFace finds faces [6]. Faces get finer lines and lighter simplification so features stay recognizable.
The AI styles use Magenta’s arbitrary image stylization network [7], which repaints your photo with the textures and colors of a reference painting.
Results are scale-aware: the preview is calculated at screen size and the download is recalculated at full resolution with the same settings, so strokes and textures keep the same look.
Three features use AI models that run on your device: subject detection, face detection and the AI neural painting styles. All 14 standard styles work without AI. If a device cannot run a model, the tool says so and keeps every other feature available.
| Feature | Model and license | Download | Needs | If unavailable |
|---|---|---|---|---|
| Subject detection | U²-Net-p, Apache 2.0 | 4.4 MB, once | WebAssembly | Paint the selection with the brush |
| Face preservation | UltraFace RFB-320, MIT | 1.3 MB, once | WebAssembly | Styles run without face-specific detail |
| AI neural painting | Magenta arbitrary image stylization, Apache 2.0 | 11.7 MB, once | WebGL (fast) or WebAssembly (slower) | Use the 14 standard styles |
Models are served from this website, cached by your browser after the first use and never send anything back. AI neural painting is generated at up to 1,024 pixels on the long side (High detail) and then scaled to your chosen download size. Pets are segmented by the subject model, but the face model is trained on human faces only.
Use a sharp, well-lit photo where the subject fills much of the frame, keep Detail on Balanced for most photos, raise it for fur and fine textures, and separate the subject from a busy background.
With this tool, yes. Your photos are processed entirely inside your web browser and are never uploaded, stored or seen by anyone else, including us. The page’s security policy blocks its scripts from connecting to any other website, so a photo could not be sent elsewhere even by mistake.
Many photo-to-cartoon apps upload your picture to a server and process it there. This converter works differently: the image engine and AI models are loaded into the page, and your photo is read, transformed and saved on your own device. Your original file is never modified; every result is a new image. Once the page and models have loaded you can disconnect from the internet and keep working.
Downloaded images contain only pixels. Camera metadata such as GPS location from your original photo is not copied into the result.
Short definitions of the techniques behind the styles.
Who makes the tool, what it does to your photo, its limits, and the research it builds on.
The converter is designed and maintained by Entertainment Couch, which publishes a collection of free online tools. It was last updated on .
Edit, blur, resize and compress images, all free in your browser with nothing to install.
Search, or pick a topic. Using the tool, styles, AI, privacy, files and more.
It is a free online tool from Entertainment Couch that turns photos into pencil sketches, cartoons, line art, comics, anime-inspired illustrations, watercolors, oil paintings and pop art. It runs in your browser, so photos are never uploaded.
Yes. Every style and feature is free, with unlimited photos, full-resolution downloads, no watermark, no sign-up and no premium version.
Add your photo, choose Pencil Sketch, adjust the intensity and detail, and press Download. The sketch is created in your browser and saved without a watermark.
Add a selfie or portrait, choose Portrait Cartoon, Cartoon or Anime-Inspired, and keep On-device AI assist turned on. It finds your face and keeps finer detail there, so the cartoon still looks like you.
Yes. Choose Watercolor, Oil Painting or Digital Illustration, or open the AI styles tab for neural paintings such as Starry Swirl and Stained Glass.
No. Open the page in any modern browser on your phone, tablet or computer and start converting. There is nothing to install and no account.
Drag it onto the drop area, press Choose photos, or copy an image and paste it with Ctrl+V on Windows or ⌘V on Mac. On phones, Choose photos opens your photo library or camera.
Press Compare and drag the slider across the image, or press and hold the Hold for original button. On a keyboard, press C to toggle compare mode.
Yes. Use the zoom buttons, the mouse wheel or a two-finger pinch, and drag to move around. The zoom value button switches between fit-to-screen and 100%.
Yes. Undo and redo cover style changes, settings and selection brush strokes. Use the buttons or Ctrl+Z and Ctrl+Shift+Z (⌘ on Mac).
Yes. Add several photos, set up the style you want, then press Apply style to all & save ZIP. Each photo is processed in turn and the results are saved together in one ZIP file.
Yes. Press Cancel in the progress box. Processing stops immediately and your previous result stays on screen.
Intensity controls how strong the effect is, such as how dark the pencil shading is or how vivid the cartoon colors are. Detail controls how much fine structure is kept: Minimal gives bold, simple shapes, while High keeps small features and texture.
The tool is designed for that. With AI assist on, faces are detected and get finer lines and lighter simplification, while the rest of the photo is stylized more strongly. Balanced or High detail keeps the most likeness.
No. It applies generic illustration techniques such as cel shading, clean line work and soft highlights to your own photo. It does not reproduce any existing character, artist or franchise.
Line Art with Detail set to Minimal gives clean, bold outlines with large areas to color. Download it as PNG for the sharpest print.
Yes. Open Subject & background and choose a background treatment: original photo, soft blur, black and white, solid color, transparent, or a different style.
Yes. The download uses the same settings and is recalculated at full resolution with strokes and textures scaled to match, so it looks like a sharper version of the preview.
Partly. The 14 standard styles use classic image-processing algorithms, which are fast and work on every device. On-device AI models add subject detection, face preservation and seven neural painting styles. We label which is which.
U²-Net-p for subject detection, UltraFace for face detection and Magenta’s arbitrary image stylization network for neural painting. All are open-source and run in your browser.
The models are files of 1.3 to 9.3 MB that run on your device. They download once from this website the first time you use an AI feature and are then cached by your browser.
The tool tells you, and all 14 standard styles, the selection brush, comparison and downloads keep working. AI painting uses WebGL when available and falls back to slower WebAssembly or CPU processing.
Yes. Choose Your style image in the AI styles tab and pick a picture. The network learns its colors and brush textures and applies them to your photo. Use artwork you have the rights to.
No. All processing, including the AI models, happens inside your browser. Your photos are never uploaded, stored or shared.
No. The tool reads your photo and creates a new image. The original file on your device is never modified.
Yes. After the page and any AI models you use have loaded, you can disconnect from the internet and keep converting photos.
No. The downloaded file is a newly created image that contains only pixels, so camera metadata such as GPS location is not copied into it.
JPG, JPEG, PNG, WebP, GIF (first frame), BMP and AVIF. HEIC photos from iPhone work in browsers that can open them, such as Safari.
PNG, JPG and WebP. PNG is lossless and keeps transparent backgrounds; JPG and WebP make smaller files with an adjustable quality setting.
Open Subject & background, set Background to Remove (transparent), check the edges with Show selection, and download as PNG or WebP.
Files can be up to 50 MB. Photos over 36 megapixels, or 16 megapixels on phones and tablets, are scaled down when opened to keep your device responsive.
Yes. Choose Original size under Download. You can also pick a smaller size for social media or messaging.
Current versions of Chrome, Edge, Firefox and Safari on Windows, macOS, Linux, Android, iPhone and iPad. JavaScript must be enabled.
Choose Minimal or Balanced detail, pick a smaller download size, close other heavy tabs, and process large batches on a computer rather than a phone.
The file may be damaged, not an image, larger than 50 MB, or in a format your browser cannot open, such as HEIC outside Safari. Convert it to JPG or PNG and try again.
Yes. Every control works with a keyboard and screen reader, the comparison slider responds to arrow keys, and animations are reduced when your device asks for reduced motion.
No questions match your search. Try a different word.