消耗细化 日志查询
总消耗查询内容在DMXAPI的位置
接口地址
https://www.dmxapi.cn/api/log/代码示例
python
import requests
from datetime import datetime
# 配置信息
URL = "https://www.dmxapi.cn/api/log/self"
# 认证信息 - 请替换为你自己的信息
SYSTEM_TOKEN = "*********************" # 替换为你的系统令牌
USER_ID = "*******" # 在 个人设置 中获得
# 每页显示条数
PAGE_SIZE = 10
# ============== 时间配置 ==============
# 查询模式选择：
# "today" - 查询今天的数据
# "yesterday" - 查询昨天的数据
# "week" - 查询最近7天的数据
# "month" - 查询最近30天的数据
# "custom" - 自定义时间范围（需要设置下方的 CUSTOM_START 和 CUSTOM_END）
QUERY_MODE = "today"
# 自定义时间范围（仅当 QUERY_MODE = "custom" 时生效）
# 格式: "YYYY-MM-DD" 或 "YYYY-MM-DD HH:MM:SS"
CUSTOM_START = "2025-01-01 00:00:00"
CUSTOM_END = "2025-01-01 23:59:59"
# =====================================
def get_log_detail(start_timestamp: int, end_timestamp: int, page: int = 1,
page_size: int = PAGE_SIZE, log_type: int = 0,
token_name: str = "", token_group: str = "",
model_name: str = "", ip: str = "",
response_id: str = "", request_id: str = ""):
"""
获取消耗明细日志数据
Args:
start_timestamp: 开始时间戳
end_timestamp: 结束时间戳
page: 页码，从1开始
page_size: 每页条数
log_type: 日志类型，0为全部，2为消耗日志
token_name: 令牌名称筛选
token_group: 令牌分组筛选
model_name: 模型名称筛选
ip: IP地址筛选
response_id: 响应ID筛选
request_id: 请求ID筛选
Returns:
dict: 包含日志数据的字典
"""
headers = {
"Accept": "application/json",
"Authorization": f"Bearer {SYSTEM_TOKEN}",
"Rix-Api-User": USER_ID,
}
params = {
"p": page,
"page_size": page_size,
"type": log_type,
"token_name": token_name,
"token_group": token_group,
"model_name": model_name,
"start_timestamp": start_timestamp,
"end_timestamp": end_timestamp,
"ip": ip,
"response_id": response_id,
"request_id": request_id
}
response = requests.get(URL, headers=headers, params=params)
if response.status_code != 200:
print(f"请求失败: {response.status_code}")
return {}
result = response.json()
if not result.get("success"):
print(f"接口返回错误: {result.get('message')}")
return {}
return result.get("data", {})
def get_all_logs(start_timestamp: int, end_timestamp: int, max_pages: int = 100):
"""
获取所有日志数据（分页获取）
Args:
start_timestamp: 开始时间戳
end_timestamp: 结束时间戳
max_pages: 最大获取页数
Returns:
list: 所有日志条目列表
"""
all_items = []
page = 1
while page <= max_pages:
data = get_log_detail(start_timestamp, end_timestamp, page=page)
if not data:
break
items = data.get("items", [])
if not items:
break
all_items.extend(items)
# 检查是否还有更多数据
total = data.get("total", 0)
if len(all_items) >= total:
break
page += 1
print(f"已获取 {len(all_items)}/{total} 条记录...")
return all_items
def print_log_detail(data: dict):
"""格式化输出日志明细"""
if not data:
print("没有数据")
return
items = data.get("items", [])
page = data.get("page", 1)
page_size = data.get("page_size", PAGE_SIZE)
total = data.get("total", 0)
print(f"分页信息: 第 {page} 页 | 每页 {page_size} 条 | 共 {total} 条记录")
print("-" * 120)
if not items:
print(" 暂无消耗记录")
return
print(f"{'序号':>4} {'模型名称':<25} {'消耗额度':>10} {'输入Token':>10} {'输出Token':>10} {'耗时(ms)':>8} {'请求时间':<20}")
print("-" * 120)
for idx, item in enumerate(items, 1):
model = item.get('model_name', 'N/A')[:23]
quota = item.get('quota', 0) / 500000
prompt_tokens = item.get('prompt_tokens', 0)
completion_tokens = item.get('completion_tokens', 0)
use_time = item.get('use_time', 0)
created_at = item.get('created_at', 0)
time_str = datetime.fromtimestamp(created_at).strftime('%Y-%m-%d %H:%M:%S') if created_at else 'N/A'
print(f"{idx:>4} {model:<25} {quota:>10.6f} {prompt_tokens:>10} {completion_tokens:>10} {use_time:>8} {time_str:<20}")
def analyze_logs(items: list) -> dict:
"""
分析日志数据
Returns:
dict: 包含分析结果的字典
"""
if not items:
return {}
# 按模型分组统计
model_stats = {}
total_prompt_tokens = 0
total_completion_tokens = 0
for item in items:
model = item.get('model_name', 'unknown')
quota = item.get('quota', 0) / 500000
prompt_tokens = item.get('prompt_tokens', 0)
completion_tokens = item.get('completion_tokens', 0)
total_prompt_tokens += prompt_tokens
total_completion_tokens += completion_tokens
if model not in model_stats:
model_stats[model] = {
'total_quota': 0,
'count': 0,
'prompt_tokens': 0,
'completion_tokens': 0
}
model_stats[model]['total_quota'] += quota
model_stats[model]['count'] += 1
model_stats[model]['prompt_tokens'] += prompt_tokens
model_stats[model]['completion_tokens'] += completion_tokens
# 计算总消耗
total_consumption = sum(stats['total_quota'] for stats in model_stats.values())
# 按消耗排序
sorted_models = sorted(model_stats.items(), key=lambda x: x[1]['total_quota'], reverse=True)
return {
'total_consumption': total_consumption,
'model_count': len(model_stats),
'request_count': len(items),
'total_prompt_tokens': total_prompt_tokens,
'total_completion_tokens': total_completion_tokens,
'model_stats': dict(sorted_models)
}
def save_analysis_report(analysis: dict, items: list, start_time: datetime, end_time: datetime, filepath: str = "消耗细化报告.txt"):
"""
保存分析报告到文件
Args:
analysis: 分析结果
items: 原始数据
start_time: 开始时间
end_time: 结束时间
filepath: 保存路径
"""
import os
# 获取脚本所在目录
script_dir = os.path.dirname(os.path.abspath(__file__))
full_path = os.path.join(script_dir, filepath)
with open(full_path, 'w', encoding='utf-8') as f:
f.write("=" * 70 + "\n")
f.write(" API 消耗细化分析报告\n")
f.write("=" * 70 + "\n\n")
f.write(f"报告生成时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n")
f.write(f"查询时间范围: {start_time.strftime('%Y-%m-%d %H:%M')} 至 {end_time.strftime('%Y-%m-%d %H:%M')}\n\n")
f.write("-" * 70 + "\n")
f.write("【总体统计】\n")
f.write("-" * 70 + "\n")
f.write(f" 总消耗额度: {analysis.get('total_consumption', 0):.4f}\n")
f.write(f" 使用模型数: {analysis.get('model_count', 0)}\n")
f.write(f" 请求总次数: {analysis.get('request_count', 0)}\n")
f.write(f" 总输入Token: {analysis.get('total_prompt_tokens', 0):,}\n")
f.write(f" 总输出Token: {analysis.get('total_completion_tokens', 0):,}\n\n")
f.write("-" * 70 + "\n")
f.write("【各模型消耗明细】\n")
f.write("-" * 70 + "\n")
model_stats = analysis.get('model_stats', {})
for i, (model, stats) in enumerate(model_stats.items(), 1):
percentage = (stats['total_quota'] / analysis['total_consumption'] * 100) if analysis['total_consumption'] > 0 else 0
f.write(f"\n {i}. {model}\n")
f.write(f" 消耗额度: {stats['total_quota']:.4f} ({percentage:.1f}%)\n")
f.write(f" 请求次数: {stats['count']}\n")
f.write(f" 输入Token: {stats['prompt_tokens']:,}\n")
f.write(f" 输出Token: {stats['completion_tokens']:,}\n")
f.write(f" 平均每次: {stats['total_quota'] / stats['count']:.6f}\n")
f.write("\n" + "-" * 70 + "\n")
f.write("【详细记录】\n")
f.write("-" * 70 + "\n")
f.write(f"{'序号':>4} {'模型名称':<25} {'额度':>10} {'输入':>8} {'输出':>8} {'时间':<20}\n")
f.write("-" * 70 + "\n")
for i, item in enumerate(items, 1):
model = item.get('model_name', 'N/A')[:23]
quota = item.get('quota', 0) / 500000
prompt = item.get('prompt_tokens', 0)
completion = item.get('completion_tokens', 0)
created_at = item.get('created_at', 0)
time_str = datetime.fromtimestamp(created_at).strftime('%Y-%m-%d %H:%M:%S') if created_at else 'N/A'
f.write(f"{i:>4} {model:<25} {quota:>10.6f} {prompt:>8} {completion:>8} {time_str:<20}\n")
f.write("\n" + "=" * 70 + "\n")
print(f"\n分析报告已保存至: {full_path}")
def get_time_range(mode: str) -> tuple:
"""
根据查询模式获取时间范围
Args:
mode: 查询模式 ("today", "yesterday", "week", "month", "custom")
Returns:
tuple: (start_timestamp, end_timestamp, start_time, end_time)
"""
from datetime import timedelta
now = datetime.now()
today = now.replace(hour=0, minute=0, second=0, microsecond=0)
if mode == "today":
start_time = today
end_time = today.replace(hour=23, minute=59, second=59)
elif mode == "yesterday":
start_time = today - timedelta(days=1)
end_time = start_time.replace(hour=23, minute=59, second=59)
elif mode == "week":
start_time = today - timedelta(days=6)
end_time = now
elif mode == "month":
start_time = today - timedelta(days=29)
end_time = now
elif mode == "custom":
try:
if len(CUSTOM_START) == 10:
start_time = datetime.strptime(CUSTOM_START, "%Y-%m-%d")
else:
start_time = datetime.strptime(CUSTOM_START, "%Y-%m-%d %H:%M:%S")
if len(CUSTOM_END) == 10:
end_time = datetime.strptime(CUSTOM_END, "%Y-%m-%d").replace(hour=23, minute=59, second=59)
else:
end_time = datetime.strptime(CUSTOM_END, "%Y-%m-%d %H:%M:%S")
except ValueError as e:
print(f"时间格式错误: {e}")
print("请使用格式: YYYY-MM-DD 或 YYYY-MM-DD HH:MM:SS")
exit(1)
else:
print(f"未知的查询模式: {mode}")
print("支持的模式: today, yesterday, week, month, custom")
exit(1)
return int(start_time.timestamp()), int(end_time.timestamp()), start_time, end_time
if __name__ == "__main__":
import time
# 根据配置获取时间范围
start_timestamp, end_timestamp, start_time, end_time = get_time_range(QUERY_MODE)
print(f"查询模式: {QUERY_MODE}")
print(f"查询时间范围: {start_time.strftime('%Y-%m-%d %H:%M:%S')} 至 {end_time.strftime('%Y-%m-%d %H:%M:%S')}")
print()
# 获取第一页数据预览
data = get_log_detail(start_timestamp, end_timestamp)
print_log_detail(data)
# 获取所有数据并分析
total = data.get("total", 0)
if total > 0:
print(f"\n正在获取全部 {total} 条记录...")
all_items = get_all_logs(start_timestamp, end_timestamp)
if all_items:
print("\n正在生成分析报告...")
analysis = analyze_logs(all_items)
print(f"\n【快速统计】")
print(f" 总消耗: {analysis['total_consumption']:.4f}")
print(f" 模型数: {analysis['model_count']}")
print(f" 请求数: {analysis['request_count']}")
print(f" 总输入Token: {analysis['total_prompt_tokens']:,}")
print(f" 总输出Token: {analysis['total_completion_tokens']:,}")
# 保存到文件
save_analysis_report(analysis, all_items, start_time, end_time)返回示例
json
查询时间范围: 2025-12-09 00:00:00 至 2025-12-09 23:59:59
分页信息: 第 1 页 | 每页 10 条 | 共 35 条记录
------------------------------------------------------------------------------------------------------------------------
序号 模型名称 消耗额度 输入Token 输出Token 耗时(ms) 请求时间
------------------------------------------------------------------------------------------------------------------------
1 gpt-5.1-chat 0.011712 10 233 6 2025-12-09 17:31:50
2 gpt-5.1-chat 0.003144 7 62 3 2025-12-09 17:06:57
3 gpt-5.1-chat 0.002994 7 59 3 2025-12-09 17:06:52
4 gpt-5.1-chat 0.003294 7 65 5 2025-12-09 16:08:50
5 gpt-5.1-chat 0.001694 7 33 3 2025-12-09 16:08:37
6 gpt-5.1-chat 0.001394 7 27 3 2025-12-09 16:08:31
7 gpt-5.1-chat 0.001494 7 29 2 2025-12-09 16:08:15
8 gpt-5.1-chat 0.002944 7 58 2 2025-12-09 16:08:11
9 gpt-5.1-chat 0.002944 7 58 3 2025-12-09 16:08:06
10 gpt-5.1-chat 0.002994 7 59 3 2025-12-09 16:07:09
正在获取全部 35 条记录...
已获取 10/35 条记录...
已获取 20/35 条记录...
已获取 30/35 条记录...
正在生成分析报告...
【快速统计】
总消耗: 0.0997
模型数: 3
请求数: 35
总输入Token: 247
总输出Token: 1,971© 2025 DMXAPI 消耗细化 日志查询 🍌