Anthropic Releases Claude Opus 4.8 With Effort Controls, Same Pricing

Anthropic released Claude Opus 4.8 today, building on Opus 4.7 with stronger benchmark performance and improved agentic capabilities. The new model is available immediately at the same price: $5 per million input tokens and $25 per million output tokens. The launch introduces effort controls on claude.ai and Cowork, allowing users to adjust how much reasoning Claude applies to responses, with higher settings producing stronger answers and lower settings prioritizing speed.

Claude Code now features dynamic workflows for handling complex assignments, including the ability to coordinate hundreds of parallel subagents. Early testers found Opus 4.8 more dependable at agentic tasks, with internal evaluations showing it is roughly four times less likely than Opus 4.7 to miss code flaws. Fast mode for Opus 4.8, operating at 2.5 times normal speed, now costs one-third of earlier versions.