Image Gen
Generate images through scripts/imagine.js, a zero-dependency Node.js CLI (Node 18+, no npm install) backed by multiple providers. It saves images to disk and prints the absolute paths.
Quick start
# Generate an image
node scripts/imagine.js "a retro synthwave poster, neon grid, 80s style" -o poster.png
# Edit an existing image
node scripts/imagine.js "add a red balloon in the sky" -e photo.png -o edited.png
# Multi-step session: keep context across calls (design first, then its three views)
node scripts/imagine.js "a cute robot mascot, front view, flat vector style" --session mascot -o design.png
node scripts/imagine.js "draw its three views (front, side, back) from the same design" --session mascot -o views.png
# Clean up the session cache when the workflow is done
node scripts/imagine.js --clear mascot
Model selection
Priority for general images: GPT Image 2 (gpt-image-2) → Nano Banana (gemini-2.5-flash-image) → SiliconFlow Qwen-Image (Qwen/Qwen-Image) → OpenAI GPT Image (gpt-image-1). When a provider has no API key, fall back to the next configured one.
| Asset |
Recommended models (priority order) |
| Images / posters |
gpt-image-2 → gemini-2.5-flash-image → Qwen/Qwen-Image → gpt-image-1 |
| Stickers / icons / logos |
Ideogram 3.0 (transparent) → gpt-image-1.5 / gpt-image-2 with --background transparent -f png → FLUX.2 pro |
| Seamless textures |
Z-Image Turbo Tiling (fal.ai) → FLUX.2 pro → any strong model with a strict tiling prompt |
Specialist models outside the script's providers (Ideogram, fal tiling) are called through their own APIs; imagine.js covers OpenAI, SiliconFlow, Gemini, and OpenAI-compatible relays.
Key workflows
- Transparent assets:
--background transparent -f png (never jpeg); phrase stickers as "die-cut sticker, thick white border, bold clean shapes".
- Seamless textures: include "seamless, tileable, top-down, flat uniform lighting, no shadows, edges wrap perfectly".
- Text in images: GPT Image 2 is best for CJK/multilingual text; keep text short and quoted.
- Aspect ratio: Gemini via
--aspect 16:9; OpenAI/SiliconFlow via -s WxH (Qwen native sizes like 1328x1328, 1664x928).
- Iterate cheaply: draft with
Qwen/Qwen-Image / Nano Banana or -q low, render finals on gpt-image-2 with -q high; --seed for reproducible output.
References — load on demand
| When |
Reference |
| Full CLI arguments, session details, exit codes |
core-usage |
| First run with no API key configured (env keys, AGENTS.md merge) |
core-setup |
| Debugging the pipeline or editing the scripts |
core-script |
1---
2name: imagine
3description: AI image, sticker, icon, and seamless-texture generation for agents. Use when the user asks to create, generate, edit, or iterate on images, photos, illustrations, posters, stickers, icons, logos, sprites, textures, or mockups. Provides a zero-dependency Node.js script that calls GPT Image 2, Nano Banana (Gemini), SiliconFlow Qwen-Image, or other image models and saves the result to disk.
4metadata:
5 author: Hairy
6 version: "2026.8.11"
7---
8
9# Image Gen
10
11Generate images through `scripts/imagine.js`, a zero-dependency Node.js CLI (Node 18+, no npm install) backed by multiple providers. It saves images to disk and prints the absolute paths.
12
13## Quick start
14
15```bash
16# Generate an image
17node scripts/imagine.js "a retro synthwave poster, neon grid, 80s style" -o poster.png
18
19# Edit an existing image
20node scripts/imagine.js "add a red balloon in the sky" -e photo.png -o edited.png
21
22# Multi-step session: keep context across calls (design first, then its three views)
23node scripts/imagine.js "a cute robot mascot, front view, flat vector style" --session mascot -o design.png
24node scripts/imagine.js "draw its three views (front, side, back) from the same design" --session mascot -o views.png
25
26# Clean up the session cache when the workflow is done
27node scripts/imagine.js --clear mascot
28```
29
30## Model selection
31
32Priority for general images: **GPT Image 2 (`gpt-image-2`) → Nano Banana (`gemini-2.5-flash-image`) → SiliconFlow Qwen-Image (`Qwen/Qwen-Image`) → OpenAI GPT Image (`gpt-image-1`)**. When a provider has no API key, fall back to the next configured one.
33
34| Asset | Recommended models (priority order) |
35|-------|-------------------------------------|
36| Images / posters | `gpt-image-2` → `gemini-2.5-flash-image` → `Qwen/Qwen-Image` → `gpt-image-1` |
37| Stickers / icons / logos | Ideogram 3.0 (transparent) → `gpt-image-1.5` / `gpt-image-2` with `--background transparent -f png` → FLUX.2 pro |
38| Seamless textures | Z-Image Turbo Tiling (fal.ai) → FLUX.2 pro → any strong model with a strict tiling prompt |
39
40Specialist models outside the script's providers (Ideogram, fal tiling) are called through their own APIs; `imagine.js` covers OpenAI, SiliconFlow, Gemini, and OpenAI-compatible relays.
41
42## Key workflows
43
44- **Transparent assets**: `--background transparent -f png` (never `jpeg`); phrase stickers as "die-cut sticker, thick white border, bold clean shapes".
45- **Seamless textures**: include "seamless, tileable, top-down, flat uniform lighting, no shadows, edges wrap perfectly".
46- **Text in images**: GPT Image 2 is best for CJK/multilingual text; keep text short and quoted.
47- **Aspect ratio**: Gemini via `--aspect 16:9`; OpenAI/SiliconFlow via `-s WxH` (Qwen native sizes like `1328x1328`, `1664x928`).
48- **Iterate cheaply**: draft with `Qwen/Qwen-Image` / Nano Banana or `-q low`, render finals on `gpt-image-2` with `-q high`; `--seed` for reproducible output.
49
50## References — load on demand
51
52| When | Reference |
53|------|-----------|
54| Full CLI arguments, session details, exit codes | [core-usage](references/core-usage.md) |
55| First run with no API key configured (env keys, AGENTS.md merge) | [core-setup](references/core-setup.md) |
56| Debugging the pipeline or editing the scripts | [core-script](references/core-script.md) |
hairyf