Microsoft Research Releases Webwright: A Terminal-Native Web Agent Framework That Scores 60.1% on Odysseys, Up from Base GPT-5.4’s 33.5%
AI Agents

Microsoft Research Releases Webwright: A Terminal-Native Web Agent Framework That Scores 60.1% on Odysseys, Up from Base GPT-5.4’s 33.5%

Asif RazzaqMarkTechPost
AI Summary

Microsoft Research released Webwright, a terminal-native web agent framework that uses Playwright scripts instead of click-trace automation. Powered by GPT-5.4, Webwright achieves 60.1% on the Odysseys benchmark and 86.7% on Online-Mind2Web, nearly doubling the base model's 33.5% performance.

This article was originally published on MarkTechPost. Read the full story at the source.

Read Full Article at MarkTechPost

Related Articles

Hermes Agent Ships Tool Search for MCP: Anthropic Evals Show 49% to 74% Accuracy Gain on Opus 4

Hermes Agent Ships Tool Search for MCP: Anthropic Evals Show 49% to 74% Accuracy Gain on Opus 4

MarkTechPost
How to Use AgentTrove: Streaming 1.7M Agentic Traces and Building a Clean ShareGPT SFT Dataset in Python

How to Use AgentTrove: Streaming 1.7M Agentic Traces and Building a Clean ShareGPT SFT Dataset in Python

MarkTechPost
Hands-On With Gemini Spark: I Gave It Access to My Life and It Friend-Zoned My Boyfriend

Hands-On With Gemini Spark: I Gave It Access to My Life and It Friend-Zoned My Boyfriend

Wired AI
So you’ve heard these AI terms and nodded along; let’s fix that

So you’ve heard these AI terms and nodded along; let’s fix that

TechCrunch AI