CONFIRMEDDate confidence: CONFIRMEDStated as firmHow firmly the source words the date: confirmed, scheduled, estimated, delayed or unverified.Show all CONFIRMED pins90% CONFIDENCEOverall confidence: 90%How well the pin's source and references back up its dates.Weighted average of how firmly 3 references, the source included, support the pin's start and end times; a reference counts half as much for every 180 days older than the newestShow all pins at 75% confidence or better
Anthropic's[1][3] own announcement, dated the day the model became available.- Start
- Jul 24, 202695% CONFIDENCEanthropic.com
Claude Opus 5 Released
Anthropic releases Claude Opus 5, a step-change improvement for the Opus tier for long-running agents, with gains in coding and professional work.
- Anthropic released Claude Opus 5 on July 24, 2026, available on all platforms as claude-opus-5 at $5/$25 per million tokens, the same as Opus 4.8[1][3]
- Positioned as close to the frontier intelligence of Claude Fable 5 at half the price; new default model on Claude Max, strongest on Claude Pro; still behind Mythos 5 on cybersecurity tasks[1]
- Cyber classifiers block binary-based vulnerability scanning, penetration testing and exploit generation; flagged requests fall back to Opus 4.8 by default; Cyber Verification Program members get a less restricted version[1]
- Beta updates: mid-conversation tool changes without invalidating the prompt cache, and automatic fallbacks on the API[1]
- The Opus product page announces it as new, "a step-change improvement for the Opus tier", available to Pro, Max, Team and Enterprise[3]
Notable features
- Opus 5 is described as a step-change improvement over Opus 4.8, with the largest gains in deep reasoning, agentic and long-horizon tasks, and test-time compute scaling; thinking is on by default and can be disabled only at effort high or below[2]
- Specifications: 1M-token context window, 128K max output tokens (300K on the Batch API in beta), adaptive thinking with high default effort, text and image input with text output, and a May 2026 reliable knowledge cutoff; Fast mode runs around 2.5 times the default speed at twice the base price[1][2]
- State of the art on Frontier-Bench and GDPval-AA; ARC-AGI 3 score three times the next-best model; Zapier AutomationBench pass rate about 1.5x next-best[1]
- Life sciences: better than Opus 4.8 on every eval, with organic chemistry up 10.2 points and protein tasks up 7.7[1]
- Alignment: called Anthropic's most aligned model to date, with a misaligned behavior score of 2.3[1]
Buy onClaude