Add GPT Image 2 image generation to your AI Agent.
GPT Image 2 • Install • API Key • HiAPI
Get API Key · Pricing · HiAPI Docs · Prompt Gallery · All HiAPI Skills
HiAPI Matrix: 🎨 Image Prompts · 🎬 Video Prompts · 🛠️ Agent Skills (you are here) · 🤖 Remote MCP · 📖 API Docs
AI Agent? Skip the README and read llms-install.md. It contains installation steps and error-handling rules written for agents.
An AI skill for OpenClaw / Claude Code / OpenCode / Codex-style agents. After installation, your AI Agent can use GPT Image 2 for image generation through HiAPI.
HiAPI is an AI API platform built for developers: one API for all AI models. Images, video, music, and text with one key.
| Skill | Description | Model |
|---|---|---|
| HiAPI GPT Image 2 | Text-to-image and image-to-image generation | GPT Image 2 family |
Need a tested starting point? Browse Awesome GPT Image 2 Prompts first. It contains output-backed visual recipes with full prompts, aspect ratios, HiAPI Draw links, and source attribution. Pick a recipe, replace the subject, product, city, brand, or copy, then use this skill to generate the adapted result.
If the generated image is meant as the starting frame of a video, plan the motion afterward with hiapi-video-prompt-generator-skill, then render with hiapi-seedance-2-0-video-skill (image-to-video).
For broader agent integration, use hiapi-skills as the directory or connect Remote MCP at https://mcp.hiapi.ai/mcp.
npx -y github:HiAPIAI/hiapi-gpt-image-2-skill -yThe installer auto-detects Codex (~/.codex/skills) and Claude Code (~/.claude/skills). If both exist, the -y flag installs to both. To target a specific agent or directory:
npx -y github:HiAPIAI/hiapi-gpt-image-2-skill --codex # ~/.codex/skills only
npx -y github:HiAPIAI/hiapi-gpt-image-2-skill --claude # ~/.claude/skills only
npx -y github:HiAPIAI/hiapi-gpt-image-2-skill --target=/path # custom directory
AGENT_SKILLS_DIR=/path npx -y github:HiAPIAI/hiapi-gpt-image-2-skill -yThe script also reports whether HIAPI_API_KEY is set and links to where to create one.
openclaw skills add https://github.com/HiAPIAI/hiapi-gpt-image-2-skillgit clone https://github.com/HiAPIAI/hiapi-gpt-image-2-skill.git
export AGENT_SKILLS_DIR="/path/to/your/agent/skills"
mkdir -p "$AGENT_SKILLS_DIR"
cp -R hiapi-gpt-image-2-skill "$AGENT_SKILLS_DIR/hiapi-gpt-image-2"Replace AGENT_SKILLS_DIR with your agent's skill directory.
Install the HiAPI GPT Image 2 image generation skill:
1. Run: npx -y github:HiAPIAI/hiapi-gpt-image-2-skill -y
(auto-detects Codex / Claude Code skill directories)
2. Set the HIAPI_API_KEY environment variable from https://www.hiapi.ai/en/dashboard/api-keys
3. Read SKILL.md for usage
- Open Get API Key
- Sign in or create a HiAPI account
- Create a new API Key
- Set the environment variable in the terminal that runs your agent:
export HIAPI_API_KEY="your_hiapi_api_key_here"
export HIAPI_BASE_URL="https://api.hiapi.ai"Check configuration:
node scripts/check-config.mjsLive check:
node scripts/check-config.mjs --liveAsk your AI Agent to generate images with natural language, or provide reference image URLs for the image-to-image variants.
- Text-to-image: describe the image you want and generate it
- Image-to-image: use
gpt-image-2/image-to-imagewith--input-urlvalues - Model variants:
gpt-image-2/text-to-image,gpt-image-2/image-to-image - Aspect ratios:
auto,1:1,3:2,2:3,4:3,3:4,5:4,4:5,16:9,9:16,2:1,1:2,3:1,1:3,21:9,9:21 - Resolutions:
1K,2K,4K - Local output: images are saved to
outputs/ - URL output: if HiAPI returns an image URL, the Agent returns the URL directly
- Clear errors: missing Key, invalid Key, insufficient balance, rate limits, and safety policy blocks all include a next step
Talk directly to your AI Agent:
Use
$hiapi-gpt-image-2to generate a 16:9 image of a sunset over the sea.
Use HiAPI GPT Image 2 to create a minimal logo, aspect ratio 1:1.
Generate a 9:16 social media poster with the headline text "Build Faster".
node scripts/hiapi-gpt-image-2.mjs \
--prompt "Create a cinematic mountain lake photo at sunset" \
--aspect-ratio 16:9Image-to-image:
node scripts/hiapi-gpt-image-2.mjs \
--model gpt-image-2/image-to-image \
--prompt "Turn this product photo into a clean premium studio ad" \
--input-url "https://example.com/product.jpg" \
--aspect-ratio 16:9 \
--resolution 2KCustom output directory:
node scripts/hiapi-gpt-image-2.mjs \
--prompt "Minimal poster for an AI image API, premium tech brand style" \
--aspect-ratio 1:1 \
--output-dir ./outputs.
├── README.md
├── README.zh-CN.md
├── SKILL.md
├── agents/
│ └── openai.yaml
├── references/
│ ├── api.md
│ └── output.md
├── scripts/
│ ├── check-config.mjs
│ ├── hiapi-gpt-image-2.mjs
│ └── lib/
│ └── gpt-image-2.mjs
├── tests/
│ └── gpt-image-2.test.mjs
└── llms-install.md
| Problem | Solution |
|---|---|
HIAPI_API_KEY is required |
Create a Key at Get API Key, then set HIAPI_API_KEY. |
401 Unauthorized |
Check whether the API Key is correct, or generate a new Key. |
402 Payment Required / insufficient balance |
Open the HiAPI Dashboard and check your account status. |
429 Too Many Requests |
Wait and retry, or reduce concurrent generation requests. |
| Content blocked | The prompt triggered a safety policy. Revise the description. |
| No image output | Check the task response; this skill expects an image URL or data URI in data.output[] after the task succeeds. |
| Skill update available | The CLI checks the HiAPI skills index at startup. If the update is optional, it prints the upgrade command and continues. |
| Skill update required | The CLI stops and prints the required upgrade command. Run npx -y github:HiAPIAI/hiapi-gpt-image-2-skill -y, then restart your agent. |
Set HIAPI_SKIP_UPDATE_CHECK=1 only for offline or locked-down environments where the skills index cannot be reached.
| Agent | Install Method |
|---|---|
| Codex | npx -y github:HiAPIAI/hiapi-gpt-image-2-skill --codex |
| Claude Code | npx -y github:HiAPIAI/hiapi-gpt-image-2-skill --claude |
| OpenClaw | openclaw skills add https://github.com/HiAPIAI/hiapi-gpt-image-2-skill |
| OpenCode | AGENT_SKILLS_DIR=~/.opencode/skills npx -y github:HiAPIAI/hiapi-gpt-image-2-skill -y |
| Cursor / other | npx -y github:HiAPIAI/hiapi-gpt-image-2-skill --target=/your/skills/dir |
MIT
HiAPI — One API, all AI models