Anthropic has released Claude Opus 4.5, marking a significant milestone in AI performance. This model is the first to breach the 80% threshold in the SWE-Bench Verified benchmark, achieving an impressive 80.9%. This surpasses the previous leader, GPT-5.1-Codex-Max, at 77.9%, and Gemini 3 Pro, which scored 76.2%.
Key Features of Claude Opus 4.5:
- Enhanced Reliability: More resistant to prompt injection attacks.
- Improved Memory Management: Stops abruptly ending lengthy conversations when reaching memory limits.
Availability: Claude Opus 4.5 is accessible to subscribers of Pro, Max, Team, and Enterprise plans.
Stay updated with the latest AI advancements!
