Gemini 3.8 Flash called a joke: "Terminal-Bench never lies"
oleks01 · x · 2026-09-08
Developer oleks01 dismisses Gemini 3.8 Flash as "a joke," arguing users shouldn't be fooled by Google's flashy comparison charts — its real Terminal-Bench results expose the gap between marketing and actual agentic coding performance.
More from Models
- Local benchmark: Flint matches Qwen3.6 quality at 50/52 while running 3.6x faster on M4 Pro — RichardSocher · 2026-10-27
- Mirendil debuts Fern, a from-scratch model with up to 10M token context built for agentic AI R&D — shayan_ · 2026-10-27
- ICML 2026 position paper: LLM 'thinking traces' are a harmful anthropomorphism — rao2z · 2026-10-27
- One stray 'u' and Claude starts spelling like a Brit — cantrell · 2026-10-27
- Cohere's Invent-A-Dataset paper: zero-seed synthetic data beats frontier APIs on quality and diversity — idanbeck · 2026-10-27
- ARC Prize redesigns leaderboard: highest reasoning effort by default, CSV export — mhmazur · 2026-10-27