以下部分引用 AI 总结，现阶段 AI 仍然有幻觉。请以内容的原文为准。

数据新鲜度群聊 07-14 ✓卖方 07-14 ✓Wrap 断档15天 ✕News 07-15 ✓X 断档12天 ✕页面生成 07-15

Agent Arena

9 atoms · 跨 1 天 · 首见 2026-06-04 · 最近 2026-06-04

三色: 🟦 fact 9 · 🟥 take 0 · stance ▲0/▼0/◆9

来源: X 9

时态: fresh:9

标签: 好数字:8 · 好信源:2

2026-06-04 · 用户任务分布
We analyzed the task distribution in Agent Arena across a 7-day window: 160K real user tasks spanning coding, debugging, research, document creation, frontend development, file analysis, and long multi-step workflows.
tags: 好数字·好信源 · → daily
2026-06-04 · 评估方法
We measure millions of live sessions where real users accomplish real tasks.
tags: 好信源 · → daily
2026-06-04 · Leaderboard 构建规模
This leaderboard snapshot is built from 300K+ tasks, 2M+ tool calls, and 40M lines of code by agents.
tags: 好数字 · → daily
value: qty=300K+ tasks, 2M+ tool calls, 40M lines of code
2026-06-04 · 任务类别占比 - Code writing
Code writing (17.5%)
tags: 好数字 · → daily
value: qty=17.5%
2026-06-04 · 任务类别占比 - Research and lookup
Research and lookup (10.8%)
tags: 好数字 · → daily
value: qty=10.8%
2026-06-04 · 任务类别占比 - Planning and brainstorming
Planning and brainstorming (10.6%)
tags: 好数字 · → daily
value: qty=10.6%
2026-06-04 · 任务类别占比 - Multimodal image/video work
Multimodal image/video work (10.2%)
tags: 好数字 · → daily
value: qty=10.2%
2026-06-04 · 任务类别占比 - Document creation
Document creation (9.1%)
tags: 好数字 · → daily
value: qty=9.1%
2026-06-04 · 任务类别占比 - Code debugging
Code debugging (8.9%)
tags: 好数字 · → daily
value: qty=8.9%