Artificial intelligence company Anthropic has introduced Claude Opus 4.8, the latest version of its flagship AI model, bringing enhanced coding performance, stronger reasoning abilities, and improved support for complex agent-based tasks. The updated model is now available through Claude.ai, Claude Code, and the Claude API under the identifier “claude-opus-4-8,” expanding access for developers and enterprise users.
A key highlight of the release is the introduction of effort controls, which allow users to decide how much computational power the model uses when generating responses. Higher effort settings consume more tokens but can deliver better results on demanding tasks. Anthropic said Claude Opus 4.8 operates with a high-effort setting by default while maintaining token usage levels similar to the previous version.
Read more : Claude Code solves year-long problem in one hour, says Google …
Despite the new capabilities, Anthropic has kept pricing unchanged from Claude Opus 4.7. The standard version costs $5 per million input tokens and $25 per million output tokens, while the faster mode costs $10 and $50 respectively. According to the company, fast mode delivers responses at roughly two and a half times the speed of the standard option.
The company has also upgraded Claude Code with dynamic workflows designed for large software development projects. This feature enables the AI assistant to plan tasks, launch parallel sub-agents, verify results, and prepare final reports. Anthropic said the system can work effectively with codebases containing hundreds of thousands of lines, making it suitable for complex enterprise environments.
Read more : Codex and Claude usage limits increased for Christmas holiday …
In addition, Anthropic introduced live update capabilities within its Messages API, allowing developers to modify instructions, permissions, token limits, and context while an AI agent continues working. These adjustments can be made without restarting tasks or interrupting prompt caching, providing greater flexibility for long-running workflows and advanced automation projects.
Anthropic claims Claude Opus 4.8 delivers noticeable improvements in coding, knowledge work, reasoning, and agentic performance compared with its predecessor. The company also reported stronger safety measures, including reduced deceptive behavior and lower compliance with misuse attempts. External evaluations from software, finance, legal, and research organizations indicated that the model completed tasks efficiently while maintaining competitive cost and performance levels.