Claude Opus 5.5 Released · Chronopin

Claude Opus 5.5 Released · Chronopin

Input and output tokens fall to $4 and $20 per million, "20% less than Opus 5" - TechCrunch and XDA give the old prices as $5 and $25 - and cache reads drop to $0.20 per million, 60% less; output also generates "more than 30% faster"^[1][3][4]
It is "our first release since we called for pacing the frontier" and was tested before release by external evaluators including Frontier Design and METR; on Anthropic's automated behavioral audit it is "the strongest-performing model we've tested to date"^[1]
Because it is "comparable to Claude Mythos 5.1 in biology and cybersecurity" it ships with safeguards like Claude Fable 5.1's, plus EU AI Act watermarking, and it is no longer available with thinking mode switched off^[1]
One early tester "completed a 680,000-line code migration in less than a day"; asked to cut load times across every page of a web app, Opus 5.5 "succeeded 39 of 40 times" where Opus 5 made smaller changes that also altered the app's behaviour^[1]
Available the same day on Amazon Web Services, Google Cloud and Microsoft Azure, and as claude-opus-5-5 on the Claude Platform; Claude Sonnet 5.5 and Claude Haiku 5.5 "will follow in the coming weeks"^[1]

Notable features

Anthropic introduces Claude Opus 5.5, "the first model in our new Claude 5.5 family", which "performs at the level of Claude Fable 5.1 on most work and costs 40% less to run than Opus 5"^[1][3]
Specifications: 1M-token context window, 128K max output tokens (300K on the Batch API in beta), adaptive thinking that is always on with medium default effort, text and image input with text output, and a June 2026 reliable knowledge cutoff^[2]
It leads Anthropic's benchmark table in agentic coding, computer use and knowledge work: Terminal-Bench 4.0 66.4% against Fable 5.1's 55.8%, Opus 5's 52.3% and GPT-6 Astra's 57.9%; GDPval-AA v2.1 1846 against 1735, 1708 and 1542^[1]
Anthropic cautions that "at these levels of capability we've found that benchmark margins have become a less reliable guide to real-world differences", and that in its own use "the gap between Opus 5.5 and Claude Fable 5.1 is narrower than these scores suggest"^[1]
TechCrunch notes the release "makes significant changes to how Opus communicates" - less jargon, important information first - which Anthropic frames as a safety benefit as well as a practical one^[1][3]