YZ Index
AI 价格追踪
主流 AI 模型 API 定价 × 评测真实成本，降价涨价一目了然
| # | 模型 | 厂商 | 输入价 $/1M tokens | 输出价 $/1M tokens | 评测实际成本 均/题 | 评测总成本 |
|---|---|---|---|---|---|---|
| 1 | DeepSeek V3 | deepseek | $0.27 | $1.10 | $0.0007 | $0.055 (80题) |
| 2 | DeepSeek R1 | deepseek | $0.55 | $2.19 | $0.0014 | $0.110 (80题) |
| 3 | Gemini 2.5 Pro | $1.25 | $10.00 | $0.0056 | $0.450 (80题) | |
| 4 | Qwen Max | alibaba | $1.60 | $6.40 | $0.0040 | $0.320 (80题) |
| 5 | GPT-4o | openai | $2.50 | $10.00 | $0.0062 | $0.500 (80题) |
| 6 | Claude Sonnet 4.6 | anthropic | $3.00 | $15.00 | $0.0090 | $0.720 (80题) |
| 7 | GPT-o3 | openai | $10.00 | $40.00 | $0.0250 | $2.000 (80题) |
| 8 | Claude Opus 4.6 | anthropic | $15.00 | $75.00 | $0.0450 | $3.600 (80题) |
价格来源于各厂商官方 API 定价页面，最后更新于配置文件。评测实际成本基于最近一次完整评测（Run #16）中每道题的 cost_est 字段计算。
价格变化箭头基于 yz_pricing_history 表自动检测，每次跑分时记录快照。