vibesbench: an informal benchmark where models opine on 90s alt-rock angst
firasd · x · 2026-10-25
Developer firasd built vibesbench, an informal benchmark that archives full transcripts of different models answering the same conversational prompts. One prompt asks why 90s American music was so angsty; in another, Claude Sonnet 4.5 delivers a thoughtful take on Courtney Love being underrated, arguing her celebrity has overshadowed her songwriting on Live Through This. Transcripts are open-sourced on GitHub.
More from Models
- AsciiBench: blind voting arena tests which AI draws better ASCII art across 13 models — JasonBotterill · 2026-10-25
- Satirical ARC-AGI-3 leaderboard shows GPT-6 near 90% score at $10K+ per task — haider1 · 2026-10-25
- Anthropic appears to be quietly nerfing Opus 5.5 usage limits, users report — CtrlAltDwayne · 2026-10-25
- Codex user reports usage down 31% after reset, calls OpenAI pricing uneconomical — CtrlAltDwayne · 2026-10-25
- Nous Research: all subscription tiers get cloud Hermes agent instances within days — Teknium · 2026-10-25
- Thinking Machines pivots from data to models and immediately hits Pareto SOTA — nirmal_dist · 2026-10-25