Your Edgerunners AI portrait looks like generic sci-fi because "Cyberpunk Edgerunners style" is a name, and an image model needs a specification: how heavy the ink is, how many tones each colour gets, what colour the shadows are, which colours dominate, and what has to stay the same as your photo. Cyber Anime Filter is a Cyberpunk Edgerunners AI filter with that specification built in for 8 styles, so you upload one photo and get 4 images without typing anything.
This post explains what that specification has to cover and why each missing piece causes a failure you have probably already seen. It does not print a prompt.
Key takeaways
- The name of the show tells a model the genre. The look only arrives when line weight, flat two-tone shading, shadow colour and palette are each spelled out.
- Likeness is a separate job from style. Unless your hair, face shape, eye colour and expression are ranked first, the model draws a good-looking stranger.
- Cyber Anime Filter runs all of this on GPT Image, an image model made by OpenAI. A set of 4 images costs 1 credit, and packs start at $9.99 for 10 credits.
What generic sci-fi looks like and why you keep getting it
You know the picture. The skin is airbrushed and softly lit. The nose and lips are modelled like a photo. There is an orange and teal glow, a blurred city behind, perhaps a visor or a glowing collar. It is a competent image. It is also nothing like a frame of Cyberpunk: Edgerunners, which is drawn with hard black lines and flat blocks of colour.
There are two reasons a short request ends up there.
The word "cyberpunk" is wide. It covers decades of films, games, book covers and concept paintings, and most of that imagery is painted or photographic. A one-line prompt leaves the model to choose what "cyberpunk" means, and it lands on the broad middle of everything the word has ever been attached to.
A photo pulls toward realism. When you attach a selfie, the model starts from an image with real skin texture, soft gradients and photographic light. Unless it is told plainly that the output is a hand-drawn animation frame, and what that means, it keeps the photo's shading and paints neon on top. The result is a photo wearing a filter.
The seven parts of the look that have to be spelled out
Cyber Anime Filter builds every image from a style sheet that was checked against frames of the series. Each row below is something the show does on purpose, next to what an under-specified request tends to return in its place.
| Part | What the Edgerunners look does | What a vague prompt returns |
|---|---|---|
| Line | Black outlines, thick on the outside of a figure and thin inside it, with corners in place of curves | Thin even lines, or no lines at all |
| Shading | Flat fills. Each area is one colour plus one hard-edged shadow | Soft gradients and airbrushed skin |
| Shadow colour | Purple or teal, never black | Grey and black |
| Palette | Three loud colours over navy in the first season. One saturated colour per character against dust in the new one | Orange and teal on everything |
| Light | A rim of colour on each side of the face. Highlights on hair and metal are hard geometric blocks | Soft bloom and glow |
| Chrome | Cyberware drawn in flat grey planes with one highlight | Mirror-finish 3D metal |
| Finish | Halftone dots and colour fringing, or scan lines, dust and a viewfinder frame | Lens blur and film grain |
Three of these do most of the work.
Line weight
The outline in this style is not a uniform stroke. It is heavy around the silhouette and light on interior detail, and it turns in corners. Faces get a small, minimal nose and large, sharp eyes. A model that is only told "anime" will reach for a softer, rounder, thinner line, because that describes far more anime than this one show does.
Flat two-tone shading
This is the rule that separates a cel from a painting. Every area of the drawing is one flat colour with one hard-edged shadow shape on it. There are no gradients and no blended skin. If this rule is missing, or if it is stated weakly, the model blends, and a blended face next to a neon sign is exactly the semi-realistic picture people complain about.
Purple and teal shadows
Shadows in this look are tinted. A cheek in shadow is purple. The dark side of a jacket is teal or blue. Black is kept for the ink. It is one of the least guessable parts of the style. A model left to its defaults shades with grey.
Why naming the show fails and describing it fails too
Here is the part that catches people who have already tried to be careful. There are two opposite ways to under-specify this style, and they fail in opposite directions.
Lean only on the name, and the model reaches for what it associates most strongly with that name, which is the cast. You asked for yourself in the style of the show and you got someone with a familiar jacket and a familiar haircut. That is the "it copied an existing character" failure.
Describe the look without anchoring it, and the description floats free. Words like "bold lines" and "flat colours" fit a thousand styles, and the model drifts back to a semi-realistic painted look with slightly harder edges.
A working set of instructions has to do both things at once and keep them in balance: point at the right reference, then spell the look out in enough detail that the model draws the style and has no reason to borrow the characters. That balance is found by testing, one render after another, on the model you are actually using. It does not transfer cleanly from one model to another.
Cyber Anime Filter also gives each of its 8 styles an original outfit and scene, so there is nothing in the request that asks for a character from the series. Our page on the Cyberpunk Edgerunners AI prompt covers what that built-in prompt has to solve.
Why your face goes missing
Style and likeness compete. Anime reduces a face to a handful of lines, so there is very little room to carry who you are. In this style, likeness comes from a short list:
- the shape and colour of your hair
- face shape and jaw
- eye colour and eyebrows
- skin tone
- expression
- the things you wear on your face: glasses, facial hair, freckles, piercings
When a request pushes hard on style and says nothing about these, the model fills them with the show's stock features. You get a sharp chin and a spiky fringe that were never yours. The fix in principle is ranking: identity has to be stated as the first priority, above the style, and the features that carry it have to be listed so the model knows what "the same person" means when the nose is a single tick of ink.
The outfit works the other way round. Cyber Anime Filter replaces your clothes on purpose, with the jacket or coat of the style you picked, and keeps your head, hair and face. That split is what lets a drawing look like a character from that world and still look like you.
One honest limit: even with identity ranked first, an AI model varies from render to render, and likeness is not guaranteed. A sharp, front-facing photo with your hair and shoulders visible gives the closest result.
Shot design: one image is a portrait and four are a character
A chatbot answers one request with one picture, usually a centred head and shoulders. The series is remembered for how it frames people: low angles, foreshortened limbs, a small figure in a huge street.
A set from Cyber Anime Filter is 4 images with 4 different jobs:
- Portrait. Chest up, three-quarter view, eyes to the camera, the scene out of focus behind.
- Action. The moment that belongs to the style, such as the after-image trail in Sandevistan Rush or the ground punch in Chrome Solo.
- Street. A full-body wide shot with the character small in the location, like an establishing shot.
- Key art. A low-angle poster pose cut out against a flat graphic background of rays and jagged shapes.
The hard part is not asking for four pictures. It is keeping them the same person. Each new request to an image model is a fresh draw, so the jacket changes colour, the hair gets longer, the cyberware moves from one arm to the other. Inside a set, all 4 images share one outfit description, one scene and one set of identity rules, so they read as one character seen by four cameras.
Format is part of shot design too. The generator asks you to choose 9:16, 1:1 or 16:9 before it draws, and each of the styles composes differently in each shape.
Why stray text and wrong hands show up
Night City is covered in signs, and signs invite lettering. Image models are happy to write, and what they write is often nonsense: half a word, a made-up logo, a signature in the corner, a border around the frame. On a fan image this is the detail that gives the game away fastest.
The principle is to remove the invitation. Cyber Anime Filter asks for no words, letters, numbers, logos or watermark anywhere in the image, and for signs and graffiti to be drawn as abstract shapes.
Hands follow the same logic. Fingers are hard for image models, and a mechanical hand is harder. Stating the count helps. It does not make errors impossible.
That is the second honest limit: AI sometimes adds stray marks or miscounts a finger no matter how the request is written. When it happens on Cyber Anime Filter, the fix is to run the set again. Our list of common filter mistakes and their fixes goes through each case.
Classic neon and the Season 2 look are two specifications
Most prompts in circulation were written for the 2022 season: night, rain, thick ink of varying weight, yellow, cyan and magenta over navy, halftone dots in the shadows and colour fringing at the edges of the frame.
Cyberpunk: Edgerunners 2 does not look like that. The new season premieres on Netflix on October 20, 2026, with 10 episodes from CD PROJEKT RED and Studio TRIGGER. Netflix's Tudum describes a reimagined animation style, and the teasers on the official site show it: bright daylight on rubble, a 1990s animation feel, dust, lens flare, handheld points of view, camcorder viewfinder frames and tape scan lines. Shadows are cool purple or blue-grey, and each lead carries one saturated colour.
A description of the first look gives you the wrong season. So Cyber Anime Filter keeps two style sheets: 4 Classic neon styles and 4 Season 2 look styles.
If you are not sure which of the new leads you sit closest to, the free Edgerunners 2 character quiz suggests a style to match.
What a built-in style gives you that a pasted prompt does not
| One-line prompt in a chatbot | Cyber Anime Filter | |
|---|---|---|
| What you provide | A photo and a sentence you keep rewording | One photo, one style, one format |
| Style rules | Whatever the sentence happened to mention | Line, shading, shadow colour, palette, light, chrome and finish, written out and tested |
| Likeness | Left to chance | Ranked first, with the features that carry it listed |
| Images per run | Usually one | 4: portrait, action, street and key art |
| Same character across images | No, each request is a fresh draw | Yes, one outfit and scene per set |
On the site you upload one clear photo, pick a style and a format, and press Draw me in. The set of 4 PNG images arrives in about a minute, with no watermark.
Where to get the look without writing anything
The Edgerunners look is a stack of decisions: weighted ink, flat fills with one hard shadow, purple and teal in the dark areas, a short palette, your identity ranked above the style, and four camera setups that agree with each other. Miss one and the picture slides back toward generic sci-fi.
That testing is what Cyber Anime Filter sells. Upload a photo to the generator, pick a style, and the set is drawn for you. A set of 4 images costs 1 credit. Credits come in one-off packs of $9.99 for 10, $24.99 for 30 and $69.99 for 100, with no free tier and no subscription, and they never expire. The same image can be animated into a 10-second silent video with Seedance 2.0 from ByteDance for 10, 22 or 40 credits at 480p, 720p or 1080p. Full details are on the pricing page.
Cyber Anime Filter is an independent fan-made tool. It is not affiliated with, endorsed by or sponsored by CD PROJEKT RED, Studio TRIGGER or Netflix, and it uses no footage, logos or character art from the series.
Frequently asked questions
Why does ChatGPT turn my Edgerunners portrait into a painting instead of anime?
A chatbot paints your Edgerunners portrait because a short request names a genre and leaves the drawing method open, which Cyber Anime Filter fixes by building the method in. Flat fills, one hard shadow per area, purple or teal shadow colour and weighted ink lines all have to be specified. Without them, an image model keeps the soft shading of the photo and adds neon on top.
Why does my AI Edgerunners picture look like David or Lucy instead of me?
An AI Edgerunners picture turns into a known character when the name of the show is the strongest signal in the request, so Cyber Anime Filter ranks your identity above the style. Its 8 styles use original outfits and scenes, and each render is told to keep your hair, face shape, eye colour and expression. Likeness still varies between renders and is not guaranteed.
Do I need to learn prompt writing to get the Edgerunners anime look?
No, Cyber Anime Filter needs no prompt writing: you upload one photo, pick one of 8 styles and a format, and press Draw me in. The style rules, the outfit, the scene and the 4 camera setups are already written and tested on GPT Image, the OpenAI model the site uses. A set of 4 images costs 1 credit.
Why do I get a different face every time I ask a chatbot for another Edgerunners image?
Each new request to an image model is a fresh draw, so the face, outfit and palette shift unless every one of them is pinned down, which Cyber Anime Filter does inside a set. All 4 images in a set share one outfit, one scene and one set of identity rules. Small differences between the 4 images remain, because AI models vary.
Can one prompt cover both the 2022 Edgerunners look and the Edgerunners 2 look?
No, Cyber Anime Filter treats them as two separate looks, with 4 Classic neon styles and 4 Season 2 look styles. The first season is night, rain and acid yellow, cyan and magenta over navy. Cyberpunk: Edgerunners 2, on Netflix from October 20, 2026, is brighter: daylight, dust, purple shadows and camcorder viewfinder frames. One description cannot be both.
Sources
- Cyberpunk: Edgerunners 2, official site (CD PROJEKT RED)
- Netflix Tudum: Cyberpunk: Edgerunners 2 release date and news