Anthropic has released Claude Opus 4.8, the latest iteration of its flagship AI model, available globally today. This update arrives just 41 days after the launch of Opus 4.7, marking a significantly accelerated development cycle for the company. The new model builds upon its predecessor with measurable improvements in coding, reasoning, and agentic capabilities, while maintaining the same pricing structure for standard usage. A primary focus of Opus 4.8 is enhanced honesty and reliability. Early testers and evaluation data indicate the model is approximately four times less likely than Opus 4.7 to overlook flaws in generated code or make unsupported claims. It demonstrates a stronger tendency to proactively flag uncertainties regarding inputs and outputs, addressing a common issue where AI models confidently assert progress without sufficient evidence. This improvement is critical for high-stakes professional workflows in legal and financial sectors. In terms of performance benchmarks, Opus 4.8 sets new records. It is the only model to complete every case end-to-end on the Super-Agent benchmark, outperforming both previous Opus versions and competitors like GPT-5.5 while maintaining cost parity. The model achieved the highest score on the Legal Agent Benchmark and scored 84% on the Online-Mind2Web test, establishing it as the strongest computer-use and browser-agent model Anthropic has tested. Additionally, the model improves tool calling efficiency, completing tasks with fewer steps and better context retention over long sessions. Anthropic also introduced several new features to accompany the release. The Dynamic Workflows tool, now in research preview for Claude Code, enables the model to tackle massive-scale problems by planning and executing hundreds of parallel subagents within a single session. This capability allows for complex operations, such as codebase migrations involving hundreds of thousands of lines of code, without human intervention. For users on claude.ai, a new effort control slider allows customization of the AI's processing intensity. Higher effort settings enable deeper reasoning for complex tasks, while lower settings provide faster responses to conserve rate limits. A Fast Mode for Opus 4.8 offers 2.5 times the speed of standard processing at a cost three times lower than previous fast modes. Furthermore, the Messages API now supports system entries within the message array, allowing developers to update instructions mid-task without breaking prompt caching. Despite these advancements, Anthropic continues to delay the general release of its most advanced models, such as the Claude Mythos Preview, due to ongoing cybersecurity safeguard development. The company expects to extend access to this higher intelligence tier to all customers in the coming weeks. The standard pricing for Opus 4.8 remains unchanged at $5 per million input tokens and $25 per million output tokens, with Fast Mode priced at $10 and $50 respectively.