The SDXL Turbo workflow is designed to streamline the process of generating high-quality images from text prompts using the SDXL model. This workflow is particularly efficient because it consolidates the image generation process into a single step, leveraging the power of the SDXL model and the Stability AI framework. At the core of this workflow are several key nodes, including the EmptyLatentImage, which initializes the latent space, and the CLIPTextEncode, which encodes the text prompt into a format that the model can understand.
The workflow utilizes the VAEDecode node to convert latent representations back into images, ensuring that the output is both detailed and visually appealing. The SamplerCustom and KSamplerSelect nodes are employed to fine-tune the sampling process, allowing for greater control over the image's style and composition. The SDTurboScheduler optimizes the generation process for speed without sacrificing quality. Finally, the SaveImage node ensures that the generated images are saved efficiently, while the MarkdownNote node provides documentation and context for each step, making this workflow both powerful and user-friendly.
API
從程式碼使用此工作流
每個 Comfy 工作流都是一個 JSON 圖。下方的 payload 就是此工作流，與 ComfyUI 執行時完全相同 — 你可以從此 URL 取得、放入版本控制，或在 ComfyUI 載入並逐節點執行。
使用 Comfy SDK 以 TypeScript 或 Python 執行。相同程式碼可用於 Comfy Cloud 或你自架的 ComfyUI — 只需更改 base URL。
// Install (beta)
npm i @comfyorg/sdk
// Run "SDXL Turbo" (TypeScript)
import { Comfy } from "@comfyorg/sdk";
const client = new Comfy({ apiKey: "comfyui-..." });
// This workflow, exported in API format (see note below)
const wf = await client.workflows.fromFile("sdxlturbo_example_api.json");
wf.setInput("7", "text", "text, watermark"); // CLIPTextEncode
const job = await client.run(wf);
await job.getOutputs("27")[0].toFile("output.png"); // SaveImageSDK 需使用 API 格式的工作流：請在 ComfyUI 開啟此工作流，並使用 檔案 → 匯出工作流（API）。
常見問題
