Introducing Claude Opus 5

Jul 24, 2026

Claude Opus 5 is available today. It’s a thoughtful and proactive model that comes close to the frontier intelligence of Claude Fable 5 at half the price.

On coding and knowledge work evaluations like Frontier-Bench and GDPval-AA, Opus 5 is the new state-of-the-art, though it remains behind Mythos 5 on cybersecurity tasks.

Opus 5 is designed to be used every day: it works more efficiently than other models. It’s the new default model on Claude Max, and the strongest model on Claude Pro.

Performance and cost-effectiveness

Claude Opus 5 provides greatly improved performance for the same cost as its predecessor, Opus 4.8. Opus 5 excels on valuable software engineering tasks. On Frontier-Bench v0.1, Opus 5 surpasses all other models, and more than doubles Opus 4.8’s performance at a lower cost per task. On CursorBench 3.2, at max effort, the model performs within 0.5% of Fable 5’s peak score, but at half the cost per task.

Related results cited by Anthropic:

• On ARC-AGI 3, Opus 5’s score is three times as high as the next-best model.
• On Zapier AutomationBench, Opus 5’s pass rate is around 1.5× the next-best model for the same cost per task.
• On OSWorld 2.0 (computer use), Opus 5 outperforms every other model at any given cost, surpassing Fable 5’s best result at just over a third of the cost.
• Life sciences: better than Opus 4.8 on every internal life-sciences eval; notably +10.2 pp on organic chemistry spectroscopy structure inference and +7.7 pp on protein-variant function prediction.

Alignment and safety

Anthropic says automated behavioral audit found Opus 5 to be its most aligned model to date (misaligned-behavior score 2.3, lowest of recent models). It adheres to Claude’s Constitution better than Opus 4.8, Sonnet 5, or Fable 5; exhibits the lowest rates of deceptive behavior; and is the least susceptible to being tricked into misuse.

Opus 5 does not advance the frontier in risky dual-use capabilities versus Mythos 5 in biology research and offensive cybersecurity. Anthropic intentionally avoided training Opus 5 on cyber tasks; it comes close to Mythos 5 at finding vulnerabilities but remains substantially behind on exploit development (OSS-Fuzz). Cyber classifiers allow source-code vulnerability finding but block binary-based vulnerability scanning, penetration testing, and exploit generation; classifiers expected to intervene ~85% less often than for Fable 5. Flagged requests fall back to Opus 4.8 by default in Claude.ai, Claude Code, and Claude Cowork.

Getting started

Claude Opus 5 is available on all platforms, priced at $5permillioninputtokensand$25 per million output tokens (same as Opus 4.8). API model ID: claude-opus-5. Fast mode (~2.5× default speed) at twice base price. Beta: mid-conversation tool changes without invalidating prompt cache; automatic fallbacks on the API when safety classifiers flag Opus 5 or Fable 5. No data retention requirements for general access (consistent with prior Opus models).

Corroboration notes (The Verge, Jul 24, 2026)

Anthropic spokesperson Danielle Ghiglieri told The Verge the company continues to work with government partners on independent testing of models including Opus 5. Marketing positions Opus 5 as the everyday/enterprise model for knowledge work and biology, with Fable 5 still preferred for the most ambitious long-horizon agent projects. Pricing framed as half of Fable 5 and slightly cheaper than OpenAI GPT-5.6.