WebArena
WebVerifiedLast updated 2026-10-04
End-to-end browser tasks in self-hosted replicas of real websites.
WebArena success rose from about 15% in 2023 to 74.3% in early 2026, leaving roughly four percentage points between the best agent and the 78.24% human baseline.
Key facts
| 2026 frontier | OpAgent 71.6% (Jan 2026); WebTactix 74.3% (Feb 2026) |
|---|---|
| Human baseline | 78.24% (original paper) |
| Credibility | Considered credible: the remaining gap to the 78.24% human baseline is a real capability gap rather than an artefact. Progress has been steep — from about 15% in 2023 to 74.3% in early 2026. |
| Verification | Verified |
| Source | WebArena leaderboard / Stanford HAI AI Index 2026 |
FAQ
What does WebArena measure?
End-to-end browser tasks in self-hosted replicas of real websites.
What is the 2026 frontier for WebArena?
OpAgent 71.6% (Jan 2026); WebTactix 74.3% (Feb 2026)
Is WebArena credible?
Considered credible: the remaining gap to the 78.24% human baseline is a real capability gap rather than an artefact. Progress has been steep — from about 15% in 2023 to 74.3% in early 2026.