Anthropic Launches Claude Opus 4.7 With Major Upgrades, But More Powerful Mythos Model Remains Under Wraps
Summary
Anthropic launches Claude Opus 4.7 with major upgrades in coding, image analysis, and creative tasks, but its most powerful model yet — Mythos — remains locked behind closed doors, available only to elite partners like Nvidia, Google, Apple, and Microsoft.
Key Points
- Anthropic releases Claude Opus 4.7, its most powerful generally available model to date, touting improvements in software engineering, image analysis, instruction-following, and creative document generation.
- Despite the release, Opus 4.7 does not advance Anthropic's capability frontier, as the previously announced Mythos Preview model outperforms it on every relevant evaluation and remains restricted to select private partners including Nvidia, JPMorgan Chase, Google, Apple, and Microsoft.
- Opus 4.7 launches with enhanced cybersecurity safeguards, serving as a testing ground for protections that Anthropic plans to eventually apply to a broader release of Mythos-class models, with a new Cyber Verification Program allowing security professionals to access relaxed restrictions for legitimate research.