# Xai Imagine > Generate high-quality images using xAI's Grok Imagine API (grok-imagine-image). Supports custom prompts, multiple aspect ratios (1:1, 16:9, 9:16, etc.), and 1k resolution. Use when the user wants to create images or visual content via xAI. Use when this capability is needed. - Skill: `tomevault-io/xai-imagine` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/xai-imagine` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/xai-imagine/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Integrations & APIs - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/xai-imagine --- # xAI Imagine Use xAI's Grok Imagine API to generate images based on text prompts. ## Quick Start To generate an image, use the `scripts/generate.js` script. It handles the API call and returns the image URL. ### Triggers - "xai 畫..." - "imagine [prompt]" - "用 grok 畫一個..." ## Aspect Ratios Supported ratios: "1:1", "3:4", "4:3", "9:16", "16:9", "2:3", "3:2", "9:19.5", "19.5:9", "9:20", "20:9", "1:2", "2:1", "auto". ## Usage Example > 需先設定 `XAI_API_KEY`： > - 建議在環境變數中設 `XAI_API_KEY` > - 或放在 `~/.openclaw/.env`（格式：`XAI_API_KEY=...`） ```bash # Generate a 16:9 image node scripts/generate.js --prompt "A futuristic neon city with a black cat" --ratio "16:9" # Equivalent (equals style) node scripts/generate.js --prompt="A futuristic neon city with a black cat" --ratio="16:9" ``` ## Workflow 1. Identify the prompt and desired aspect ratio (default is 1:1). 2. Execute `scripts/generate.js` with the `--prompt` and optional `--ratio`. 3. The script will output a JSON with the image URL. 4. Use the `message` tool to deliver the image to the user. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/nitekat1124) — claim your Tome and manage your conversions.