WebArena

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

WebArena

Why it matters

In an evaluation on real, working websites, a GPT-4 agent scored 14.41% versus 78.24% for humans, nailing down the gap between demos and reality

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/2307.13854

Added in Week 35, 2026 · See that week

On the map — ③ Tools & Agents line

  1. Previous stopfunction calling
  2. This record · stopWebArena
  3. Next stopGAIA benchmark
See it on the map

Same week

Skeleton-of-Thought publishedPaperRT-2Paper

Nearby on the same axis

MetaGPTPaperAutoGenPaperOpenAI launches function callingOfficialDevDayOfficial