Gemini Pro runs research task for nearly 5 hours with barely any progress

teortaxesTex · x · 2026-09-08

User teortaxesTex shares a real-world observation: a Pro-tier Gemini task without Deep Research has been running for almost 5 hours and has barely made any progress. The original work was done with Gemini 6 Pro; the long-running instance is just continuing the discussion.

His real question: how much can an Astra-tier model automate the pruning of research directions — helping researchers weed out dead-end technical paths. Despite the slow progress, he remains bullish on AI-assisted research automation.

Related event: Users Test Marathon AI Agent Tasks on Gemini and OpenAI With Little Progress(2 posts)→

Original post →

More from Models

Models channel →