Anthropic ships Claude Sonnet 5.5: 30% faster, half of Opus 5.5's price, 70.6% on Terminal-Bench 4.0
机器之心 · wechat · 2026-09-29
Anthropic released Claude Sonnet 5.5, over 30% faster and up to 30% cheaper than Sonnet 5, complementing Opus 5.5 for everyday tasks. Highlights:
- Terminal-Bench 4.0 agent coding: 70.6% vs Sonnet 5's 10.3%
- FrontierCode: +10 points over Sonnet 5 at 1/15 per-task cost; within 2 points of Opus 5.5 on CursorBench and GDPval-AA
- First Sonnet to beat Pokémon Red from screenshots alone
Priced at half of Opus 5.5 (cache reads $0.20/M tokens); at low effort it beats Sonnet 5's best at 1/10 the cost. Safety: audited across 1,850 scenarios, high-risk cyber tasks fall back to Sonnet 5, and it's the first Sonnet shipped with a classifier defending against distillation attacks plus extended preserved thinking.
Related event: Anthropic Launches Claude Sonnet 5.5: Over 30% Faster, Up to 30% Cheaper(58 posts)→
More from Models
- Jason Calacanis threatens to ditch $20k/yr Claude as $40/mo Grok Bot outperforms it — chrisfleck · 2026-11-13
- Dan Shipper's weekend plan: let Claude Fable 5.2 build a custom Supernote OS — danshipper · 2026-11-13
- Alibaba teases Qwen4 with stronger coding and agentic capabilities at Nov 17 Hong Kong summit — teortaxesTex · 2026-11-13
- Baidu opens ERNIE 5.5 Flash Preview via OpenRouter — Baidu_Inc · 2026-11-13
- Hybrid Decision Model learns when to reason: 72.5% lower latency at higher accuracy — hbouammar · 2026-11-13
- AI×Bio benchmarks called out as LLM-generated "benchslop" that may be impairing bio-reasoning — anshulkundaje · 2026-11-13