Grok Imagine Image 2.0: Text to Image
Generate images from text prompts using Grok Imagine Image 2.0, an Aurora-engine model that excels at crisp short-text rendering and 1K or 2K outputs across 14 aspect ratios. Enter one text prompt and receive one generated image, with low or medium quality options to control cost. Ideal for poster and cover design, product and brand visuals, and character-driven scenes where consistency matters.
简介
This ComfyUI workflow turns a single text prompt into a finished image using Grok Imagine Image 2.0, an Aurora‑engine model optimized for crisp short‑text rendering and high‑quality 1K or 2K outputs across 14 aspect ratios. The GrokImageNode handles the API call to the Grok model, and the SaveImageAdvanced node writes the result to disk with your chosen filename pattern and location.
你需要提供
- 正在读取工作流
在 ComfyUI 中运行。 此工作流无法在本页面运行。你可以在 Comfy Cloud 中打开(已预先加载所需内容),或下载后在自己的 ComfyUI 中使用。
在 Cloud 中试用常见问答
Use low for quick, low-cost drafts and concept exploration. Switch to medium for improved detail and more reliable text legibility. Medium may take longer and cost more, but it’s a better choice for final outputs.
Keep prompts concise and include the exact text you want (use quotes if helpful). Specify typography cues like 'clean, bold type' and ensure a simple, high-contrast background. Avoid very long strings; short phrases render more cleanly.
The GrokImageNode exposes 14 preset aspect ratios and supports 1K or 2K outputs. Pick an aspect ratio that matches the destination (e.g., square for profiles, wide for banners). Choose 2K for higher detail or print-ready work, and 1K for web or fast iteration.
Reuse the same core prompt and descriptors. Keep character attributes stable (e.g., clothing, colors, scene). If your GrokImageNode build exposes a seed field, set and reuse the same seed between runs to reduce variation.