gpt-image

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass, no license.

Audit 0.4.1 · source sha 4599ef28a2a7 · 28 Oct 2026Report an issue or request a re-scan

• Publisherqu-skills GitHub organizationOwner of the repository this skill is read from
• Source folderqu-skills/skills/tools/image/gpt-imageThe exact folder we audited
• LicenceNot declaredAs the repository states it
• Commit pinned for this verdict4599ef28a2a7A new audit runs when the source changes
• Repository statusNot archivedLast push 12 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: Generate and edit images with OpenAI GPT-Image-2.5 and GPT-Image-2 via inference.sh CLI. Models: GPT-Image-2.5 Flare, GPT-Image-2.5 Sunburst, GPT-Image-2. Capabilities: text-to-image, image editing, inpainting,…

---
name: gpt-image
description: "Generate and edit images with OpenAI GPT-Image-2.5 and GPT-Image-2 via inference.sh CLI. Models: GPT-Image-2.5 Flare, GPT-Image-2.5 Sunburst, GPT-Image-2. Capabilities: text-to-image, image editing, inpainting, mask-based editing, multi-image reference, batch generation, transparent backgrounds. Use for: product mockups, marketing visuals, image editing, concept art, inpainting, photo manipulation. Triggers: gpt image, gpt-image-2.5, gpt image 2.5, flare, sunburst, gpt-image-2, openai image, chatgpt image, dall-e, dalle, openai image generation, gpt image edit, gpt inpainting, openai dall-e, gpt 4o image, transparent background, transparent png, remove background"
allowed-tools: Bash(belt *)
---

> **Install the belt CLI skill:** `npx skills add belt-sh/cli`

# GPT-Image

Generate and edit images with OpenAI's GPT-Image-2.5 and GPT-Image-2 via [inference.sh](https://inference.sh) CLI.

## Models

| Model | App ID | Pick it for |
|-------|--------|-------------|
| **GPT-Image-2.5 Flare** | `openai/gpt-image-2-5-flare` | Default. Higher quality than GPT-Image-2 at 50% lower latency |
| **GPT-Image-2.5 Sunburst** | `openai/gpt-image-2-5-sunburst` | Premium edits: keeps subject and composition, tighter control across multi-turn edits |
| GPT-Image-2 | `openai/gpt-image-2` | Previous generation |

All three share the same input schema. The 2.5 models add `xhigh` and `max` quality tiers.

## Quick Start

> Requires inference.sh CLI (`belt`). [Install instructions](https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md)

```bash
belt login

belt app run openai/gpt-image-2-5-flare --input '{"prompt": "a cat astronaut floating in space"}'
```


## Capabilities

Every GPT-Image app supports text-to-image generation, image editing with reference images, and mask-based inpainting through a single app.

| Feature | Description |
|---------|-------------|
| Text-to-Image | Generate images from text prompts |
| Image Editing | Edit images using reference images |
| Inpainting | Mask-based editing of specific regions |
| Batch Generation | Generate up to 10 images at once |
| Multiple Formats | PNG, JPEG, WebP output |
| Transparent Background | Alpha-channel PNG/WebP output for stickers, icons, product cutouts |
| Flexible Resolution | Any size in 16px increments (256–3840), aspect ratio up to 3:1 |

## Examples

### Text-to-Image

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "professional product photo of sneakers on a white background, studio lighting",
  "quality": "high"
}'
```

### Multiple Images

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "minimalist logo design for a coffee shop",
  "n": 4,
  "quality": "medium"
}'
```

### Image Editing with Reference

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "change the background to a beach at sunset",
  "images": ["https://your-image.jpg"]
}'
```

### Precise Edit with Sunburst

Sunburst is built to change only what you ask for and preserve the rest.

```bash
belt app run openai/gpt-image-2-5-sunburst --input '{
  "prompt": "change the jacket to red leather, keep the person, pose and background unchanged",
  "images": ["https://your-photo.jpg"],
  "quality": "high"
}'
```

### Maximum Detail

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "macro photo of a dragonfly wing, intricate veins, morning dew",
  "quality": "max"
}'
```

### Multi-Image Reference

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "combine these two characters into one scene",
  "images": ["https://character1.jpg", "https://character2.jpg"]
}'
```

### Inpainting with Mask

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "replace with a red sports car",
  "images": ["https://street-scene.jpg"],
  "mask": "https://car-mask.png"
}'
```

### Transparent Background

Prompt for an isolated subject — describing a scene or backdrop makes the model draw one. Requires `png` (default) or `webp` output.

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "a single red apple with a green leaf, isolated subject",
  "background": "transparent"
}'
```

### Custom Resolution

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "wide cinematic landscape, mountains at golden hour",
  "width": 1920,
  "height": 1080,
  "quality": "high"
}'
```

### Fast Drafts

```bash
belt app run openai/gpt-image-2-5-flare --input '{
  "prompt": "quick concept sketch of a robot",
  "quality": "low"
}'
```

## Pricing

Approximate price per 1024x1024 image. Flare and Sunburst cost the same.

| Quality | GPT-Image-2.5 | GPT-Image-2 |
|---------|---------------|-------------|
| low | $0.006 | $0.006 |
| medium | $0.013 | $0.053 |
| high | $0.053 | $0.21 |
| xhigh | $0.094 | – |
| max | $0.21 | – |

Larger resolutions cost more. Edits add about $0.008 per reference image. See `belt app get openai/gpt-image-2-5-flare` for full pricing details.

## Parameters

| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| `prompt` | string | required | Text prompt describing the image |
| `images` | array | - | Reference image(s) for editing |
| `mask` | string | - | Mask image for inpainting |
| `n` | integer | 1 | Number of images (1–10) |
| `quality` | string | auto | low, medium, high, xhigh, max (xhigh/max are 2.5 only) |
| `width` | integer | 1024 | Output width (256–3840, multiples of 16, ratio ≤ 3:1) |
| `height` | integer | 1024 | Output height (256–3840, multiples of 16, ratio ≤ 3:1) |
| `output_format` | string | png | png, jpeg, or webp |
| `output_compression` | integer | - | Compression level for jpeg/webp (0–100) |
| `background` | string | auto | auto, transparent, or opaque (transparent needs png/webp) |

## Related Skills

```bash
# Full platform skill (all apps)
npx skills add inference-sh/skills@infsh-cli

# All image genera

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

npx skills add inference-sh/skills@infsh-cli
npx skills add inference-sh/skills@ai-image-generation
npx skills add inference-sh/skills@flux-image
npx skills add inference-sh/skills@p-image

git clone https://github.com/qu-skills/skills
git -C skills checkout 4599ef28a2a7

Works in

1. 2026-10-28
   ✓SAFE B 89 / 100

   first audit · source 4599ef28a2a7

Questions about gpt-image

• github.com/qu-skills/skills/tree/main/tools/image/gpt-image — the audited source at commit 4599ef28a2a7 and every finding aboveObserved · 28 Oct 2026

Spotted something wrong? Report a correction · Methodology