Claude Opus 5.5 Matches Fable 5.1 At 40% Less Than Opus 5
0:00 / 0:51
News
Claude Opus 5.5 Matches Fable 5.1 At 40% Less Than Opus 5
calendar_today Date:
schedule Duration: 0:51
database
Summary Report
Opus 5.5 matches Fable 5.1 on most tasks, leads GPT-6 Astra on agentic coding and knowledge work, and costs 40% less to run than Opus 5.
- 01. Terminal-Bench 4.0 66.4% vs Fable 5.1 55.8% and Astra 57.9%; CursorBench 57.8%; GDPval-AA 1846; HLE 67.7%
- 02. $4 in and $20 out; fast mode 2.5x for $8/$40; 1M context; higher subscription limits and a bankable rate-limit reset
Anthropic has released Claude Opus 5.5, the first of a new Claude 5.5 family. It performs at Claude Fable 5.1 level on most tasks and on Anthropic's table leads Fable and GPT-6 Astra on Terminal-Bench 4.0 (66.4%), CursorBench 4.0 (57.8%), GDPval-AA (1846) and Humanity's Last Exam (67.7%); Astra keeps AutomationBench and Terminal-Bench-Science. Pricing drops to $4 in and $20 out per million tokens, about 40% less than Opus 5 on typical workloads, with a 2.5x fast mode, a million-token context and 30% faster output. Subscribers get higher five-hour limits on Pro, Max and Team plus a rate-limit reset they can bank.
Meta Data
LLM:
