Skip to content
Hot storiesContinuously updated

评测显示Claude在Agent场景下任务成本仅为Gemini一半

1 reports1 reporting sourcesUpdated 14 hours ago

Start here

AI overview

2026年11月2日，一项针对生产级数据分析智能体的实测结果显示，Claude 组合（Haiku 5.5 + Sonnet 5.5）与 Gemini 组合（3.5 Flash Lite + 3.8 Flash）在成本表现上呈现明显差异。尽管 Claude 模型的单 token 标价更高，但其完成单轮任务的实际成本为 0.067 美元，相比之下 Gemini 组合单轮成本为 0.139 美元，Claude 在实际 Agent 任务中的单次回答成本仅为 Gemini 的一半左右。

AI-generated from reports · Updated 13 hours ago

Latest developmentsNov 2 23:31
生产级智能体实测显示Claude组合单轮任务成本（$0.067）约为Gemini（$0.139）的一半。

Report timeline

Follow the reports to see different sides of the story.

Nov 2
  1. Hacker News · AISelected
    Benchmark Shows Claude vs. Gemini in Production Agents: Claude Costs More Per Token but Cuts Per-Task Cost in Half

    In real-world benchmarks of production-grade data analysis agents, the Claude pair (Haiku 5.5 + Sonnet 5.5)—despite higher per-token list prices—halved the cost of completing single-turn tasks compared to the Gemini pair (3.5 Flash Lite + 3.8 Flash), coming in at $0.067 versus $0.139.

Story heat trend

Current heat 7·Comparable peak 10（Nov 3 01:00）·Comparable change, last 24 hours–

02.557.510Nov3Nov3Nov3Nov3

Trends compare only the same parties observed continuously, so the scope can be narrower than current heat counts. Move the pointer or tap the chart to see hourly heat; use the left and right arrow keys on a keyboard.