Better-Gemini

Better-Gemini
★ 0

集成 Better-GeminiGoogle GeminixAI Grok图像生成
在 ComfyUI 中调用 Google Gemini 与 xAI Grok 进行图像生成的增强节点，支持最新 v3 API 与简化的密钥配置与安装流程。
💡 在 ComfyUI 流程中用 Gemini/Grok 快速生成图像并管理 API 密钥
🍴 1 Forks💻 Python🔄 2026-03-25
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/2df45d172dc1
📦 requirements.txt
google-genai>=0.5.0
Better Gemini example workflow
📄 README

ComfyUI Better Gemini

Custom ComfyUI node(s) for generating images with Google Gemini and xAI Grok.

Install

  • Clone into your ComfyUI custom_nodes directory:
  • cd /path/to/ComfyUI/custom_nodes
  • git clone https://github.com//ComfyUI-Better-Gemini.git
  • Install Python deps with uv (network required):
  • uv sync
  • uv pip install --python /path/to/ComfyUI/python/bin/python -e ./ComfyUI-Better-Gemini
  • Set API keys as needed:
  • export GOOGLE_API_KEY="..." (or GEMINI_API_KEY) for Gemini
  • export XAI_API_KEY="..." for Grok
  • Restart ComfyUI.

    • Nodes

  • Better Gemini (image)
  • Inputs: prompt, model (dropdown; populated via models.list()), prompt_images (optional), aspect ratio, resolution / width+height, temperature, top_p/top_k, max tokens, thinking difficulty, seed.
  • Outputs: IMAGE, STRING (any returned text / notes).
  • Better Grok (image)
  • Inputs: prompt, model (dropdown; populated via /v1/image-generation-models when XAI_API_KEY is available), prompt_images (optional), aspect ratio, resolution, n.
  • Outputs: IMAGE, STRING (status / notes).

    • Example Workflow

    Notes

  • The Gemini node imports google-genai lazily so ComfyUI can still boot even if dependencies aren’t installed yet; execution will raise a clear error until installed.
  • This extension uses ComfyUI’s V3 extension loader (comfy_entrypoint).
  • The Gemini model dropdown is populated via client.models.list() (filtered to models supporting generateContent). It requires an API key via GOOGLE_API_KEY/GEMINI_API_KEY; otherwise it falls back to bundled defaults and logs a warning: gemini-3.1-flash-image-preview, gemini-3-pro-image-preview, imagen-4.0-generate-001, imagen-4.0-ultra-generate-001.
  • The Grok model dropdown is populated via xAI’s /v1/image-generation-models. It requires XAI_API_KEY; otherwise it falls back to bundled defaults and logs a warning: grok-imagine-image, grok-imagine-image-pro.
  • Gemini requires seed to fit in an int32; larger ComfyUI seeds are deterministically folded via seed % 2**31.
  • Grok image generation is wired against xAI’s documented image endpoints and requests response_format="b64_json", so the node can return image tensors directly instead of downloading temporary URLs.
  • The Grok HTTP client sends an explicit application User-Agent because api.x.ai can reject the default Python-urllib signature with Cloudflare 1010.
  • Grok image edits use xAI’s JSON-based /v1/images/edits API and send ComfyUI IMAGE inputs as PNG data URIs. Multiple prompt images are supported for edit/merge workflows.
  • resolution/aspect_ratio are best-effort (model-dependent). The node logs a warning if the returned size doesn’t match (no auto-resize).
  • If Gemini or Grok returns no images, the node returns a blank placeholder image and includes a note in the STRING output.

    • Dev

  • Sync dev environment: uv sync --dev
  • Run unit tests: uv run python -m unittest discover -s tests -p 'test_*.py' -v