灵感搜索工具

API 端点

POST https://jimeng.jianying.com/mweb/v1/get_explore

Python 示例

import requests
import hashlib
import time

def generate_sign(uri_path):
    device_time = int(time.time())
    sign_str = f"9e2c|{uri_path[-7:]}|7|5.8.0|{device_time}||11ac"
    sign = hashlib.md5(sign_str.encode()).hexdigest()
    return sign, device_time

def inspiration_search(sessionid, keyword="", category_id=11222, offset=0, count=20):
    uri = "/mweb/v1/get_explore"
    sign, device_time = generate_sign(uri)
    
    headers = {
        "Content-Type": "application/json",
        "Appid": "513695",
        "Appvr": "5.8.0",
        "Pf": "7",
        "Origin": "https://jimeng.jianying.com",
        "Referer": "https://jimeng.jianying.com",
        "Cookie": f"sessionid={sessionid}",
        "Device-Time": str(device_time),
        "Sign": sign,
        "Sign-Ver": "1"
    }
    
    data = {
        "count": count,
        "offset": offset,
        "filter": {
            "work_type_list": ["video", "image", "canvas"]
        },
        "category_id": category_id,
        "feed_refer": "feed_enterauto",
        "image_info": {
            "width": 2048,
            "height": 2048,
            "format": "webp",
            "image_scene_list": [
                {"scene": "smart_crop", "width": 720, "height": 720, "format": "webp", "uniq_key": "smart_crop-w:720-h:720"},
                {"scene": "loss", "width": 720, "height": 720, "format": "webp", "uniq_key": "720"}
            ]
        }
    }
    
    if keyword:
        data["keyword"] = keyword
    
    resp = requests.post(
        f"https://jimeng.jianying.com{uri}",
        params={"aid": 513695, "device_platform": "web", "region": "CN", "da_version": "3.3.8"},
        headers=headers,
        json=data
    )
    return resp.json()

def search_guess(sessionid):
    uri = "/mweb/search/v1/guess"
    sign, device_time = generate_sign(uri)
    
    headers = {
        "Content-Type": "application/json",
        "Appid": "513695",
        "Appvr": "5.8.0",
        "Pf": "7",
        "Origin": "https://jimeng.jianying.com",
        "Referer": "https://jimeng.jianying.com",
        "Cookie": f"sessionid={sessionid}",
        "Device-Time": str(device_time),
        "Sign": sign,
        "Sign-Ver": "1"
    }
    
    resp = requests.post(
        f"https://jimeng.jianying.com{uri}",
        params={"aid": 513695},
        headers=headers,
        json={"search_channel": "inspiration"}
    )
    return resp.json()

参数说明

字段	类型	必填	说明
keyword	string	否	搜索关键词 (为空时返回推荐内容)
category_id	int	否	分类ID，默认 11222
offset	int	否	结果偏移量，用于翻页
count	int	否	返回数量，默认 20
filter.work_type_list	array	否	作品类型过滤 ["video", "image", "canvas"]

搜索建议 API

/mweb/search/v1/guess 返回热门搜索词：

{
  "data": {
    "guess_list": [
      {"gid": "0", "word": "毛笔字"},
      {"gid": "0", "word": "ip"},
      {"gid": "0", "word": "海报制作"}
    ]
  }
}

Query 构造规则

有图片时

提取：[主体] + [风格] + [核心元素]

• 错误：类似的图片
• 正确：摩登上海 陆家嘴 蓝色调 海报

纯文本时

直接提取关键词

Offset 计算规则

条件 A → offset = 上轮offset + count

• Query 与上轮语义相同
• 用户说"再找找"、"换一批"、"更多"
• 新图与上一张图属于同一主题

条件 B → offset = 0

• Query 语义完全改变
• 用户说"重新开始"、"重置"

响应结构

{
  "ret": "0",
  "data": {
    "has_more": true,
    "next_offset": 20,
    "item_list": [
      {
        "common_attr": {
          "id": "xxx",
          "title": "作品标题",
          "cover_url": "https://...",
          "cover_url_map": {"720": "https://..."}
        },
        "author": {
          "name": "作者名",
          "avatar_url": "https://..."
        },
        "aigc_image_params": {
          "text2image_params": {
            "prompt": "生成时使用的提示词"
          }
        }
      }
    ]
  }
}

注意

• 连续搜索时严禁重置 offset 为 0（除非条件B）
• 关键词重叠度 >80% 时必须递增 offset
• 使用 next_offset 作为下一页的 offset 值