{
  "usage": {
    "prompt_tokens": 34,
    "completion_tokens": 89,
    "total_tokens": 123
  }
}

{
  "usage": {
    "prompt_tokens": 1520,
    "completion_tokens": 85,
    "total_tokens": 1605,
    "prompt_tokens_details": {
      "cached_tokens": 1480,
      "cache_creation_input_tokens": 0
    }
  }
}

{
  "usage": {
    "prompt_tokens": 50,
    "completion_tokens": 300,
    "total_tokens": 350,
    "completion_tokens_details": {
      "reasoning_tokens": 245
    }
  }
}

# pip install tiktoken
import tiktoken

# 使用 Qwen 分词器
encoding = tiktoken.get_encoding("o200k_base")
tokens = encoding.encode("Your prompt text here")
print(f"Token count: {len(tokens)}")

image_tokens = ceil(height / 28) × ceil(width / 28) / 4 + 2

import math

def estimate_image_tokens(width, height, max_pixels=1003520, min_pixels=3136):
  """估算 Qwen 视觉模型的图片 Token 数量。"""
  # 缩放至像素预算范围内
  total_pixels = width * height
  if total_pixels > max_pixels:
    scale = math.sqrt(max_pixels / total_pixels)
    width = int(width * scale)
    height = int(height * scale)

  # 取整到 28 的最近倍数
  width = max(28, round(width / 28) * 28)
  height = max(28, round(height / 28) * 28)

  # 计算 Token 数
  return (height // 28) * (width // 28) // 4 + 2

# 示例
print(estimate_image_tokens(1024, 1024))   # ~326 tokens
print(estimate_image_tokens(1920, 1080))   # ~326 tokens
print(estimate_image_tokens(4096, 4096))   # ~326 tokens（缩放后）

cost = (input_tokens × 输入单价) + (output_tokens × 输出单价)