Anthropic Launches Claude Opus 4.8 With Enhanced Performance and New Features
  • News
  • North America

Anthropic Launches Claude Opus 4.8 With Enhanced Performance and New Features

The new model boasts improved judgment, agentic skills, and introduces dynamic workflows

5/29/2026
Yassine Benadou
Back to News

Anthropic has officially launched Claude Opus 4.8, the latest iteration of its flagship artificial intelligence model. This new version promises significant improvements in reasoning, coding, and agentic capabilities, positioning it as a more effective and reliable collaborator for complex tasks. The upgrade is available immediately to all users at the same price point as its predecessor, accompanied by several new platform features.


Enhanced Performance and Reliability

Claude Opus 4.8 demonstrates marked advancements across a range of industry benchmarks, outperforming previous versions in coding, reasoning, and practical knowledge work. The model exhibits sharper judgment and greater reliability, making it more suitable for sophisticated, multi-step agentic tasks. Early evaluations confirm its superior performance compared to both its predecessor and other leading models in the market.

A key improvement highlighted by Anthropic is the model's increased honesty and self-awareness during task execution. Opus 4.8 is reportedly more inclined to flag uncertainties and avoid making unsupported claims, a crucial trait for enterprise applications. This is evidenced by internal tests showing it is approximately four times less likely to let flaws in its own code pass unremarked.

Strong Endorsement from Industry Partners

The new model has received positive feedback from early testers across various sectors, including legal, finance, and software engineering. Partners such as Databricks and Devin have praised its ability to handle complex, autonomous workflows with greater consistency and efficiency. This industry validation underscores the model's readiness for real-world, high-stakes professional environments where accuracy is paramount.

In specialized fields, Opus 4.8 has set new performance records, scoring highest on the Legal Agent Benchmark and excelling on tests for computer and browser-based agent tasks. Companies like CoCounsel and Hebbia noted meaningful improvements in reasoning quality and citation precision for legal and financial document analysis. These endorsements suggest the model's enhancements translate directly into tangible benefits for professional users.

New Features for Greater User Control

Alongside the model upgrade, Anthropic introduced several new features designed to enhance user experience and expand capabilities. A new "Effort Control" setting allows users to balance response quality against speed and cost, providing greater flexibility for different use cases. This feature gives users direct influence over the model's resource allocation for a given task.

For developers and enterprise clients, the introduction of "Dynamic Workflows" in Claude Code represents a significant step forward. This feature enables the model to plan and execute large-scale tasks by running hundreds of parallel subagents, suitable for codebase-wide migrations. Additionally, the model's "fast mode" is now three times cheaper, making high-speed operations more accessible.

Pricing, Availability, and Future Roadmap

Claude Opus 4.8 is available to developers today through the Claude API, with pricing for standard usage remaining unchanged from the previous version. The cost is set at $5 per million input tokens and $25 per million output tokens, ensuring continuity for existing customers. The reduced price for fast mode further enhances the model's value proposition for time-sensitive applications.

Looking ahead, Anthropic has signaled its intention to release more powerful and cost-effective models in the near future. The company is actively developing Mythos-class models, currently in limited preview for cybersecurity work under Project Glasswing. A broader release of these next-generation AI systems is expected in the coming weeks, promising even higher levels of intelligence.


The launch of Claude Opus 4.8 marks a deliberate and impactful step in Anthropic's strategy, emphasizing iterative improvement in reliability and practical utility for enterprise users. By pairing a more capable model with enhanced user controls and a clear vision for the future, the company reinforces its competitive standing. This release provides a tangible upgrade for current users while setting the stage for more significant advancements on the horizon.