AI Pixel-Art / 2D Art Tools
Companion to tools/SKILL.md (hand-authoring DCCs:
Aseprite, Tiled, LDtk, etc.). This file focuses on AI-driven generation.
Decision in 30 seconds
| Goal |
Pick |
| Pixel art with consistent palette + animations |
PixelLab (purpose-built) or Retro Diffusion (local) |
| Many style-locked assets for a single project |
Scenario (custom-trained model per project) |
| One-off concept art, mood boards, painterly 2D |
Midjourney or Leonardo PhotoReal |
| Full control + free + offline |
Stable Diffusion + pixel-art LoRA |
| Generate a tileset / icon set |
PixelLab (tilemap mode) or Scenario (with tile prompt) |
| In-engine, no external pipeline |
Scenario or Leonardo (both have Unity / Unreal SDKs) |
The first instinct should usually be: placeholder = AI, final = human.
AI is great to unblock prototyping; final shipped art usually still
benefits from a human pass for cohesion.
The contenders
PixelLab (pixellab.ai)
Purpose-built for pixel art. Models trained on pixel sprites, not natural
images, so output is genuinely pixel-perfect (no fake "pixel filter on a
photo" artifacts).
- Strengths: native sprite-sheet output, animation generation
(skeleton + interpolation), rotation / pose control, palette
selection, tilemap mode, Aseprite plugin (generate inside the DCC).
- API: REST, returns PNG + sidecar JSON (palette, pose data).
- Pricing: subscription, credits per generation. Free tier limited.
- License: commercial use OK on paid tiers, generated assets are
yours.
- Weakness: smaller stylistic range than general models; opinionated
output style.
Scenario (scenario.com)
Train a custom model on your project's reference art. Lock the style
across thousands of generations.
- Strengths: per-project style consistency, ControlNet-like
prompts (sketch → asset, depth → asset), batch generation, in-app
editor, tile-able mode, isometric mode.
- API: REST + Unity/Unreal SDKs.
- Pricing: subscription tiers, GPU-hours model.
- License: commercial use, full ownership of generated assets.
- Sweet spot: studios shipping a game with a specific visual identity
who'd otherwise burn weeks on manual asset variation.
Leonardo.AI
General-purpose image generator with strong game-asset focus.
- Strengths: PixelArt model built-in, image-to-image, tile-able
toggle, large library of community fine-tunes, decent free tier.
- Weakness: pixel art is one mode among many — not as dialled in as
PixelLab.
- API: REST, async job model.
- Pricing: token-based subscription.
- License: Pro tier required for commercial use of generated assets.
Retro Diffusion (retrodiffusion.ai)
Stable Diffusion fine-tune specialized in pixel art. Runs locally if
you have the GPU; also offered as a hosted service.
- Strengths: full control (you own the weights), zero per-image cost
if running locally, plays nice with ComfyUI / Automatic1111 workflows,
Aseprite extension.
- Weakness: needs a competent GPU (8 GB+ VRAM) for local; setup is
not turn-key.
- License: model licensed per their terms (check for commercial
shipping); generated images are yours.
Stable Diffusion + LoRAs (DIY)
Bring your own model. Common pixel-art LoRAs:
All-In-One-Pixel-Model, Pixel Art XL, PixelLandscape,
M_Pixel_Style.
- Strengths: free, total control, integrate into any pipeline
(ComfyUI graph nodes), composable with ControlNet, IP-Adapter, etc.
- Weakness: steepest learning curve; results vary widely with
prompting; need to manage models / versions / seeds yourself.
- Best for: technical artists who want a programmable pipeline, or
cost-sensitive solo devs.
General-purpose (Midjourney, DALL-E 3, Imagen 3)
Not pixel-art specialists. Useful for:
- Concept art, mood boards, palette inspiration.
- Higher-resolution painterly 2D (UI illustrations, key art, marketing).
- "Pixel art–style" output is hit-or-miss; usually ends up as
pixel-filtered raster, not true pixel art.
Don't ship sprites straight out of Midjourney unless the visual identity
intentionally embraces that look (e.g. some indie titles use it for
backdrops, not characters).
Comparison matrix
| Tool |
Pixel-true |
Animation |
Tile-able |
Custom style |
Local run |
Commercial OK |
Sweet spot |
| PixelLab |
YES |
YES (skeleton interp) |
tileset mode |
preset styles |
NO |
paid tier |
sprites + animations |
| Scenario |
depends on model |
NO (static) |
YES |
YES (per-project) |
NO |
YES |
style-locked asset libraries |
| Leonardo |
mode |
NO |
YES |
community LoRAs |
NO |
Pro tier |
mixed 2D + concept |
| Retro Diffusion |
YES |
NO |
depends |
yes via finetune |
YES |
check license |
free local pixel art |
| SD + LoRA (DIY) |
depends |
with AnimateDiff |
tileable LoRAs |
full |
YES |
per LoRA license |
pipeline builders |
| Midjourney |
NO (raster filter) |
NO |
mode |
style ref |
NO |
Pro tier |
concept art / key art |
Integration patterns
Aseprite plugin (PixelLab, Retro Diffusion)
Generate directly inside the DCC. Output lands as a new layer or sprite
sheet ready to clean up. Best for indies — minimal context switching.
REST API → engine pipeline (Scenario, Leonardo, PixelLab)
Build an editor-time tool inside Unity/Godot that calls the API on
demand, drops the result into the project asset folder. Best for
studios producing many variants (skins, recolors, palette swaps).
ComfyUI graph (SD + LoRA + ControlNet)
Programmable node graph: input sketch → ControlNet → LoRA → upscale →
posterize. Runs locally. Best for technical artists who want
reproducible pipelines.
Batch script (Python + API)
For asset libraries (icons, portraits, items): script generates N
variants, hand-pick the best. Common for card games, RPG item sets.
Hybrid workflows that work
Pure-AI rarely ships untouched. The patterns that actually work:
- AI for variants, human for hero assets: hand-paint the player
character, AI-generate 50 NPCs in the same style.
- AI generates → human cleans: faster than from-scratch, more cohesive
than raw AI.
- AI for palette / mood, human for sprites: use Midjourney to lock the
palette and atmosphere, hand-paint the sprites.
- AI for placeholder, human for shipped: prototype with AI, replace
with hand-pixel as the design firms up.
Cost reality check (May 2026 pricing)
| Tool |
Entry tier |
Per-image (typical) |
Notes |
| PixelLab |
$10–20/mo |
~$0.05 |
Subscription with credit pool |
| Scenario |
$20–80/mo |
~$0.02–0.10 |
GPU-hour metered |
| Leonardo |
$10–30/mo |
~$0.01–0.05 |
Token system |
| Retro Diffusion (hosted) |
$5–20/mo |
~$0.02 |
Local = free |
| Stable Diffusion local |
$0 |
$0 |
Hardware + electricity |
| Midjourney |
$10–60/mo |
~$0.02 |
GPU-hour metered |
For a 100-asset library: ~$2–10 with most paid tools, ~$0 local. For a
1000-asset library: paid tools become real costs; local SD wins.
Anti-patterns
| Anti-Pattern |
Why it bites |
Better |
| Ship raw AI sprites without cleanup |
Inconsistency, weird artifacts, inhuman silhouettes |
Always at least a polish pass |
| Mix outputs from 3+ different models |
Style drift, palette chaos |
Lock to one tool / style |
| Generate at high res then downscale to "pixel art" |
Anti-aliasing artifacts; not real pixel art |
Use a pixel-art-specific tool |
| Trust the "tile-able" toggle on general models |
Often visible seams |
Test with a 4×4 grid before committing |
| Forget the commercial license |
Free tier ≠ commercial use on most platforms |
Read the ToS for every tool you ship from |
| Animate by generating per-frame independently |
No frame-to-frame coherence; flicker |
Use animation-aware tools (PixelLab, AnimateDiff) or interpolate |
| Use AI for hero / promotional art |
Marketing art is judged harshly; AI shows |
Hand-do the cover, key art, store icons |
See also
1---2name: gamedev-2d-art-ai-art-tools3description: AI-assisted 2D / pixel-art generation tools and services. Covers pixel-art-specific generators (PixelLab, Retro Diffusion), game-asset generators (Scenario, Leonardo.AI), DIY Stable Diffusion + LoRAs, and general models with pixel-art prompting (Midjourney, DALL-E, Imagen). Trade-off framework: control vs. speed, palette consistency, animation coherence, tile-ability, commercial license, runs locally vs. cloud, cost per asset / per month. USE WHEN: deciding which AI tool fits a project, generating placeholder assets, generating final assets at scale, integrating an AI sprite pipeline, evaluating commercial license, comparing API costs. DO NOT USE FOR: hand-painted-only workflows (use `tools` skill), 3D / model generation (out of scope), engine integration of finished sprites (use `tools` + engine-specific skill).4---56# AI Pixel-Art / 2D Art Tools78> Companion to [`tools/SKILL.md`](../tools/SKILL.md) (hand-authoring DCCs:9> Aseprite, Tiled, LDtk, etc.). This file focuses on AI-driven generation.1011## Decision in 30 seconds1213| Goal | Pick |14|------|------|15| Pixel art with consistent palette + animations | **PixelLab** (purpose-built) or **Retro Diffusion** (local) |16| Many style-locked assets for a single project | **Scenario** (custom-trained model per project) |17| One-off concept art, mood boards, painterly 2D | **Midjourney** or **Leonardo PhotoReal** |18| Full control + free + offline | **Stable Diffusion + pixel-art LoRA** |19| Generate a tileset / icon set | **PixelLab** (tilemap mode) or **Scenario** (with tile prompt) |20| In-engine, no external pipeline | **Scenario** or **Leonardo** (both have Unity / Unreal SDKs) |2122The first instinct should usually be: **placeholder = AI, final = human**.23AI is great to unblock prototyping; final shipped art usually still24benefits from a human pass for cohesion.2526## The contenders2728### PixelLab (pixellab.ai)2930Purpose-built for pixel art. Models trained on pixel sprites, not natural31images, so output is genuinely pixel-perfect (no fake "pixel filter on a32photo" artifacts).3334- **Strengths**: native sprite-sheet output, **animation generation**35 (skeleton + interpolation), **rotation / pose control**, palette36 selection, tilemap mode, **Aseprite plugin** (generate inside the DCC).37- **API**: REST, returns PNG + sidecar JSON (palette, pose data).38- **Pricing**: subscription, credits per generation. Free tier limited.39- **License**: commercial use OK on paid tiers, generated assets are40 yours.41- **Weakness**: smaller stylistic range than general models; opinionated42 output style.4344### Scenario (scenario.com)4546Train a custom model on **your** project's reference art. Lock the style47across thousands of generations.4849- **Strengths**: per-project style consistency, **ControlNet-like**50 prompts (sketch → asset, depth → asset), batch generation, in-app51 editor, **tile-able mode**, **isometric mode**.52- **API**: REST + Unity/Unreal SDKs.53- **Pricing**: subscription tiers, GPU-hours model.54- **License**: commercial use, full ownership of generated assets.55- **Sweet spot**: studios shipping a game with a specific visual identity56 who'd otherwise burn weeks on manual asset variation.5758### Leonardo.AI5960General-purpose image generator with strong game-asset focus.6162- **Strengths**: **PixelArt model** built-in, image-to-image, **tile-able63 toggle**, large library of community fine-tunes, decent free tier.64- **Weakness**: pixel art is one mode among many — not as dialled in as65 PixelLab.66- **API**: REST, async job model.67- **Pricing**: token-based subscription.68- **License**: Pro tier required for commercial use of generated assets.6970### Retro Diffusion (retrodiffusion.ai)7172Stable Diffusion fine-tune specialized in pixel art. Runs **locally** if73you have the GPU; also offered as a hosted service.7475- **Strengths**: full control (you own the weights), zero per-image cost76 if running locally, plays nice with ComfyUI / Automatic1111 workflows,77 **Aseprite extension**.78- **Weakness**: needs a competent GPU (8 GB+ VRAM) for local; setup is79 not turn-key.80- **License**: model licensed per their terms (check for commercial81 shipping); generated images are yours.8283### Stable Diffusion + LoRAs (DIY)8485Bring your own model. Common pixel-art LoRAs:86**All-In-One-Pixel-Model**, **Pixel Art XL**, **PixelLandscape**,87**M_Pixel_Style**.8889- **Strengths**: free, total control, integrate into any pipeline90 (ComfyUI graph nodes), composable with ControlNet, IP-Adapter, etc.91- **Weakness**: steepest learning curve; results vary widely with92 prompting; need to manage models / versions / seeds yourself.93- **Best for**: technical artists who want a programmable pipeline, or94 cost-sensitive solo devs.9596### General-purpose (Midjourney, DALL-E 3, Imagen 3)9798Not pixel-art specialists. Useful for:99- Concept art, mood boards, palette inspiration.100- Higher-resolution painterly 2D (UI illustrations, key art, marketing).101- "Pixel art–style" output is hit-or-miss; usually ends up as102 pixel-filtered raster, not true pixel art.103104Don't ship sprites straight out of Midjourney unless the visual identity105intentionally embraces that look (e.g. some indie titles use it for106backdrops, not characters).107108## Comparison matrix109110| Tool | Pixel-true | Animation | Tile-able | Custom style | Local run | Commercial OK | Sweet spot |111|------|-----------|-----------|-----------|--------------|-----------|---------------|------------|112| **PixelLab** | YES | YES (skeleton interp) | tileset mode | preset styles | NO | paid tier | sprites + animations |113| **Scenario** | depends on model | NO (static) | YES | YES (per-project) | NO | YES | style-locked asset libraries |114| **Leonardo** | mode | NO | YES | community LoRAs | NO | Pro tier | mixed 2D + concept |115| **Retro Diffusion** | YES | NO | depends | yes via finetune | YES | check license | free local pixel art |116| **SD + LoRA (DIY)** | depends | with AnimateDiff | tileable LoRAs | full | YES | per LoRA license | pipeline builders |117| **Midjourney** | NO (raster filter) | NO | mode | style ref | NO | Pro tier | concept art / key art |118119## Integration patterns120121### Aseprite plugin (PixelLab, Retro Diffusion)122Generate directly inside the DCC. Output lands as a new layer or sprite123sheet ready to clean up. **Best for indies** — minimal context switching.124125### REST API → engine pipeline (Scenario, Leonardo, PixelLab)126Build an editor-time tool inside Unity/Godot that calls the API on127demand, drops the result into the project asset folder. **Best for128studios** producing many variants (skins, recolors, palette swaps).129130### ComfyUI graph (SD + LoRA + ControlNet)131Programmable node graph: input sketch → ControlNet → LoRA → upscale →132posterize. Runs locally. **Best for technical artists** who want133reproducible pipelines.134135### Batch script (Python + API)136For asset libraries (icons, portraits, items): script generates N137variants, hand-pick the best. Common for card games, RPG item sets.138139## Hybrid workflows that work140141Pure-AI rarely ships untouched. The patterns that actually work:142143- **AI for variants, human for hero assets**: hand-paint the player144 character, AI-generate 50 NPCs in the same style.145- **AI generates → human cleans**: faster than from-scratch, more cohesive146 than raw AI.147- **AI for palette / mood, human for sprites**: use Midjourney to lock the148 palette and atmosphere, hand-paint the sprites.149- **AI for placeholder, human for shipped**: prototype with AI, replace150 with hand-pixel as the design firms up.151152## Cost reality check (May 2026 pricing)153154| Tool | Entry tier | Per-image (typical) | Notes |155|------|------------|--------------------|---------------------|156| PixelLab | $10–20/mo | ~$0.05 | Subscription with credit pool |157| Scenario | $20–80/mo | ~$0.02–0.10 | GPU-hour metered |158| Leonardo | $10–30/mo | ~$0.01–0.05 | Token system |159| Retro Diffusion (hosted) | $5–20/mo | ~$0.02 | Local = free |160| Stable Diffusion local | $0 | $0 | Hardware + electricity |161| Midjourney | $10–60/mo | ~$0.02 | GPU-hour metered |162163For a 100-asset library: ~$2–10 with most paid tools, ~$0 local. For a1641000-asset library: paid tools become real costs; local SD wins.165166## Anti-patterns167168| Anti-Pattern | Why it bites | Better |169|--------------|--------------|--------|170| Ship raw AI sprites without cleanup | Inconsistency, weird artifacts, inhuman silhouettes | Always at least a polish pass |171| Mix outputs from 3+ different models | Style drift, palette chaos | Lock to one tool / style |172| Generate at high res then downscale to "pixel art" | Anti-aliasing artifacts; not real pixel art | Use a pixel-art-specific tool |173| Trust the "tile-able" toggle on general models | Often visible seams | Test with a 4×4 grid before committing |174| Forget the commercial license | Free tier ≠ commercial use on most platforms | Read the ToS for every tool you ship from |175| Animate by generating per-frame independently | No frame-to-frame coherence; flicker | Use animation-aware tools (PixelLab, AnimateDiff) or interpolate |176| Use AI for hero / promotional art | Marketing art is judged harshly; AI shows | Hand-do the cover, key art, store icons |177178## See also179180- [tools/SKILL.md](../tools/SKILL.md) — hand-authoring DCCs (Aseprite, Tiled, LDtk)181- [palettes/SKILL.md](../palettes/SKILL.md) — palette discipline (lock the palette before AI generation)182- [pixel-art-fundamentals/SKILL.md](../pixel-art-fundamentals/SKILL.md) — what counts as "real" pixel art (so you can tell when AI gets it wrong)183- PixelLab API docs: https://www.pixellab.ai/docs184- Scenario docs: https://docs.scenario.com/185- Retro Diffusion: https://retrodiffusion.ai/186- Civitai (LoRA library): https://civitai.com/
claude-dev-suite