Sonnet 5.5 Is Not a Very Good Model, Says Founder Bindu Reddy
bindureddy · x · 2026-09-29
Bindu Reddy calls Sonnet 5.5 "not a very good model": it spins in max mode, underperforms on agentic coding, and is slightly worse than Terra. He recommends sticking with Sonnet 4.6 or DeepSeek Flash, and argues Anthropic should retire both Sonnet and Haiku lines and ship their big next-gen Model-2 instead.
More from Models
- Every's vibe check: Sonnet 5.5 is 30% faster and up to 30% cheaper than Sonnet 5 — every · 2026-09-29
- Sonnet 5.5's lower price makes token-heavy agent workflows affordable, says Every — every · 2026-09-29
- Opus 5.5 sets Bug Hunt Bench record with 800+ self-check turns — PawelHuryn · 2026-09-29
- Tip: don't use Sonnet 5.5 at max — it's dumber and far more expensive; high hits the best cost per task — airesearch12 · 2026-09-29
- Vibe Check: Sonnet 5.5 Is a More Capable Partner Than Sonnet 5, If You Keep a Hand on the Wheel — kieranklaassen · 2026-09-29
- Every team tests Sonnet 5.5: half Opus's token price, matches it on writing readability — kieranklaassen · 2026-09-29