Claude Opus 4.8
3 hours ago
- #Claude Opus
- #Anthropic
- #AI Models
- Anthropic has released Claude Opus 4.8, an upgraded version of Opus 4.7 with improvements across benchmarks and enhanced collaboration capabilities.
- Opus 4.8 introduces new features including user-controlled effort levels on claude.ai, dynamic workflows in Claude Code for large-scale problems, and a cheaper fast mode that operates at 2.5x speed.
- The model shows improved reliability, judgment, and efficiency in agentic tasks, coding, legal work, and enterprise applications, according to early testers.
- Opus 4.8 demonstrates better honesty by flagging uncertainties and is less likely to make unsupported claims compared to its predecessor.
- Alongside Opus 4.8, updates include dynamic workflows in Claude Code for Enterprise, Team, and Max plans, effort control in claude.ai and Cowork, and enhanced Messages API functionality.
- Anthropic plans to release more cost-effective models and a higher-intelligence Mythos-class model in the coming weeks, pending cyber safeguards.
- Claude Opus 4.8 is available globally at the same pricing as Opus 4.7 for regular usage, with reduced pricing for fast mode.