Developer: no model comes close to Gemini when it comes to hallucinations
prasenx · x · 2026-11-12
Developer Prasenjit posted a pointed one-liner claiming that when it comes to hallucinations, no model comes close to Gemini — a sarcastic jab implying Gemini stands out for the wrong reason. No benchmarks or specifics were provided.
More from Models
- ChatGPT Pro User Finds 30% of Overnight Usage Went to Agents Self-Teaching Drawing — amplifiedamp · 2026-11-12
- LLMs are still terrible at information presentation, even when their analysis is good — rickasaurus · 2026-11-12
- Every: Fable 5.2 built a production video editor overnight—choosing the pricier model can cut total task cost — every · 2026-11-12
- Dev says ChatCut's AI edit points got much worse after switching from Claude to Gemini — yihui_indie · 2026-11-12
- Dev picks gpt-oss as his favorite modern transformer architecture — qtnx_ · 2026-11-12
- User tests suggest ChatGPT routes some chats to newer 6.2 Sol with June 13 cutoff — SuperAngryGuy · 2026-11-12