Claude Opus 4.5
2 hours ago
- #Anthropic
- #AI
- #Claude Opus
- Claude Opus 4.5 is the newest AI model from Anthropic, offering state-of-the-art performance in coding, agents, and computer use.
- It excels in deep research, working with slides and spreadsheets, and complex multi-step reasoning tasks.
- Opus 4.5 is more efficient, using fewer tokens to solve problems, and is priced at $5/$25 per million tokens.
- The model shows significant improvements in coding benchmarks, refactoring, and autonomous task execution.
- It outperforms previous models in long-horizon tasks, reducing token usage by up to 65% while maintaining quality.
- Opus 4.5 is robust against prompt injection attacks and is considered the most aligned frontier model to date.
- New features include an 'effort parameter' for API users, allowing control over token usage and capability tradeoffs.
- Product updates include Claude Code with Plan Mode, desktop app integration, and expanded access to Claude for Excel.
- The model demonstrates creative problem-solving, such as finding legitimate workarounds in airline booking scenarios.
- Anthropic has partnered with Microsoft and NVIDIA, scaling Claude on Azure with significant compute capacity investments.