AI·Img
目录 · API 文档

POST /v1/images/edits

图生图，OpenAI images.edit 兼容。multipart 上传 1-3 张参考图。

请求

  • Method: POST
  • Content-Type: multipart/form-data
  • Auth: Authorization: Bearer aiimg-sk-live-...

表单字段

字段类型必填说明
prompttext是编辑指令（如 "把背景换成雪地"）
image 或 image[]file × 1-3是png / jpeg / webp / gif，单张 ≤ 50MB
modeltext否默认 gpt-image-2
nint否1-4，默认 1
sizetext否同 generations 端点
qualitytext否low/medium/high/auto
response_formattext否url 或 b64_json

响应(200)

同 /v1/images/generations。

示例

curl

curl https://aiimg.top/v1/images/edits \
  -H "Authorization: Bearer $AIIMG_API_KEY" \
  -F "model=gpt-image-2" \
  -F "prompt=把背景换成雪地，主体保留" \
  -F "image=@./input.png" \
  -F "n=1" \
  -F "quality=low"

Python(OpenAI SDK)

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ["AIIMG_API_KEY"],
    base_url="https://aiimg.top/v1",
)

with open("input.png", "rb") as f:
    resp = client.images.edit(
        model="gpt-image-2",
        image=f,
        prompt="把背景换成雪地，主体保留",
        n=1,
        size="1024x1024",
        quality="low",
    )
print(resp.data[0].url)

Node(OpenAI SDK)

import OpenAI, { toFile } from "openai";
import fs from "node:fs";

const client = new OpenAI({
  apiKey: process.env.AIIMG_API_KEY,
  baseURL: "https://aiimg.top/v1",
});

const r = await client.images.edit({
  model: "gpt-image-2",
  image: await toFile(fs.createReadStream("input.png"), "input.png"),
  prompt: "把背景换成雪地，主体保留",
  n: 1,
  size: "1024x1024",
  quality: "low",
});
console.log(r.data[0].url);

多张参考图

curl https://aiimg.top/v1/images/edits \
  -H "Authorization: Bearer $AIIMG_API_KEY" \
  -F "prompt=融合这两张图的元素" \
  -F "image[]=@./a.png" \
  -F "image[]=@./b.png"