Nano Banana 系列开发指南

curl --location 'https://api.apiyi.com/v1beta/models/gemini-3-pro-image-preview:generateContent' \
  --header 'Authorization: Bearer sk-' \
  --header 'Content-Type: application/json' \
  --data '{
      "contents": [
          {
              "parts": [
                  {
                      "fileData": {
                          "fileUri": "https://raw.githubusercontent.com/apiyi-api/ai-api-code-samples/refs/heads/main/Vision-API-OpenAI/otter.png",
                          "mimeType": "image/png"
                      }
                  },
                  {
                      "text": "add five dogs"
                  }
              ],
              "role": "user"
          }
      ],
      "generationConfig": {"responseModalities": ["IMAGE"],
      "imageConfig": {
        "aspectRatio": "16:9",
        "imageSize": "2K"
      }},
      "safetySettings": []
  }'   > output.json

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Gemini 3 Pro Image - 图片编辑（file_uri 最小化版本）
用途：仅用于快速验证接口可用性
"""

import requests
import base64
import json
from pathlib import Path
from datetime import datetime

# ============================================================================
# 配置区域
# ============================================================================

API_KEY = "sk-"
API_URL = "https://api.apiyi.com/v1beta/models/gemini-3-pro-image-preview:generateContent"

# 图片 URL
IMAGE_URL = "https://raw.githubusercontent.com/apiyi-api/ai-pics/refs/heads/main/1762260696217_dd0352c1f9604540.png"
IMAGE_MIME_TYPE = "image/png"

# 编辑指令
EDIT_PROMPT = "将照片中的人的衣服换成蓝色夹克，头发换成紫色渐变色，人物的动作、眼睛朝向等其他结构不变"
SYSTEM_PROMPT = "您是一位专业的图像描述和生成专家。您的任务是根据用户的请求，创作出细节丰富、艺术风格明确的高质量图像提示，或对现有图像进行准确、有创意的编辑。"

# 输出参数
ASPECT_RATIO = "9:16"
RESOLUTION = "4K"
MAX_OUTPUT_TOKENS = 8000
OUTPUT_FILE = f"minimal_{datetime.now().strftime('%Y%m%d_%H%M%S')}.png"

# ============================================================================
# 核心代码
# ============================================================================

def main():
    print("=" * 60)
    print("开始测试 file_uri 格式接口")
    print("=" * 60)
    print(f"图片 URL: {IMAGE_URL[:80]}...")
    print(f"编辑指令: {EDIT_PROMPT}")
    print(f"输出参数: {RESOLUTION}, {ASPECT_RATIO}")
    print("-" * 60)

    # 构建请求体
    # 注意：fileData、mimeType、fileUri 必须使用驼峰命名
    payload = {
        "generationConfig": {
            "responseModalities": ["IMAGE", "TEXT"],
            "imageConfig": {
                "imageSize": RESOLUTION,
                "aspectRatio": ASPECT_RATIO
            },
            "maxOutputTokens": MAX_OUTPUT_TOKENS
        },
        "contents": [
            {
                "role": "model",
                "parts": [{"text": SYSTEM_PROMPT}]
            },
            {
                "role": "user",
                "parts": [
                    {
                        "fileData": {           # 驼峰命名：fileData（不是 file_data）
                            "mimeType": IMAGE_MIME_TYPE,  # 驼峰命名：mimeType
                            "fileUri": IMAGE_URL          # 驼峰命名：fileUri
                        }
                    },
                    {"text": EDIT_PROMPT}
                ]
            }
        ]
    }

    # 发送请求
    print("\n正在发送请求...")
    try:
        response = requests.post(
            API_URL,
            json=payload,
            headers={
                "Content-Type": "application/json",
                "Authorization": f"Bearer {API_KEY}"
            },
            timeout=300
        )

        print(f"响应状态码: {response.status_code}")

        if response.status_code != 200:
            print(f"❌ 错误: {response.text}")
            return

        # 解析响应
        data = response.json()
        print("✅ 成功获取响应")

        # 保存完整响应（方便调试）
        with open(OUTPUT_FILE + ".response.json", "w", encoding="utf-8") as f:
            json.dump(data, f, indent=2, ensure_ascii=False)
        print(f"📄 响应已保存: {OUTPUT_FILE}.response.json")

        # 提取并打印文本
        parts = data["candidates"][0]["content"]["parts"]
        for part in parts:
            if "text" in part:
                print(f"\n💬 文本响应: {part['text']}")

        # 保存图片
        for part in parts:
            if "inlineData" in part or "inline_data" in part:
                image_data = part.get("inlineData", part.get("inline_data", {})).get("data")
                if image_data:
                    image_bytes = base64.b64decode(image_data)
                    with open(OUTPUT_FILE, "wb") as f:
                        f.write(image_bytes)
                    print(f"\n✅ 图片已保存: {OUTPUT_FILE}")
                    print(f"📦 文件大小: {len(image_bytes) / 1024:.1f} KB")
                    print(f"🔗 文件路径: {Path(OUTPUT_FILE).resolve()}")
                    return

        print("⚠️  响应中未找到图片数据")

    except requests.Timeout:
        print("❌ 请求超时")
    except Exception as e:
        print(f"❌ 错误: {e}")

if __name__ == "__main__":
    main()
    print("\n" + "=" * 60)
    print("测试结束")
    print("=" * 60)

API Connection Error: HTTPSConnectionPool(host='api.apiyi.com', port=443): Read timed out. (read timeout=120)

timeout = {
    "1K": 300,  # 5 分钟 - 快速预览
    "2K": 300,  # 5 分钟 - 推荐使用
    "4K": 600,  # 10 分钟 - 超高清
}