Claude Opus 4.7 Launches with Enhanced Capabilities and Token Usage
- Anthropic released Claude Opus 4.7, claiming it as the most capable Opus model yet.
- Opus 4.7 scored significantly higher on benchmarks, with an increase from Opus 4.6’s scores of up to over a thousand Elo points in some tests.
- The model features visible chain-of-thought reasoning and consumes more tokens than its predecessor.
- Enhanced safeguards are included for cybersecurity, limiting certain high-risk requests.
Claude Opus 4.7 demonstrates significant improvements in coding and reasoning capabilities compared to its predecessor, addressing previous user concerns about performance degradation in version 4.6. However, its increased token usage may require users to adjust their plans or budgets accordingly.
Despite these advancements, cybersecurity features remain deliberately limited to ensure safety before wider deployment of more powerful models like Claude Mythos Preview remains restricted to vetted security firms, ensuring that Anthropic’s safety measures are robust enough for future releases.
