Introducing Claude Opus 4.8 Anthropic has released Claude Opus 4.8, an upgraded version of its AI model, offering improved performance on benchmarks and enhanced collaboration capabilities at the same price. The update includes new features such as dynamic workflows for complex coding tasks and effort control for users on claude.ai. Early testers report noticeable improvements in reliability, judgment, and efficiency across a range of applications, including coding, legal work, and general reasoning.

• Claude Opus 4.8 is now available, offering improvements over Opus 4.7 across benchmarks.
• New features include ‘dynamic workflows’ for Claude Code and ‘effort control’ for claude.ai users.
• Opus 4.8 demonstrates enhanced reliability, judgment, and efficiency in agentic tasks and coding.
• The model achieves top scores on benchmarks for legal work, computer use, and browser agents.
• A key improvement is increased honesty, with Opus 4.8 being less likely to make unsupported claims.
• Alignment assessments show Opus 4.8 has lower rates of misaligned behavior than Opus 4.7.
• Future developments include models with similar capabilities at lower costs and a new class of even more intelligent models like Claude Mythos Preview. Continue reading https://foxvector.com/articles/4cb49171-42f5-4dcb-a87f-402b42d7cf2d