Anthropic releases Opus 5.5 with lower prices and Fable-level performance
Anthropic drops Opus 5.5’s price by 20% and outperforms Fable in coding tests, while staying safe‑guarded.
Why Now
Anthropic released Opus 5.5 on September 22, 2026, two months after Opus 5, announcing cheaper rates and higher performance.
What Happened
Opus 5.5 charges $20 per million output tokens versus $25 for Opus 5, and runs faster due to lower compute. It beats the larger Fable model on many benchmarks and passes informal tasks that Fable failed. The model is less jargon‑heavy and front‑loads key information. Safety safeguards mirror those of Fable, limiting exploit discovery and biological weapon design.
Why It Matters
Lower cost and faster inference make high‑capability LLMs more accessible for coding and knowledge work, potentially accelerating adoption in enterprise tools. The performance leap also raises the bar for competitors, while the maintained safeguards signal Anthropic’s commitment to responsible scaling.
The Limitation
The performance claims are based on Anthropic’s own benchmarks; independent third‑party testing is not yet available.
What You Can Do
Try Opus 5.5 for a coding‑heavy project and compare latency and cost against your current LLM.
Source
Read original sourceWhy we picked this
Anthropic’s Opus 5.5 release is core AI content.