Anthropic Releases Claude Opus 4.8 with Cost Reductions and Enhanced Coding Features
Anthropic launched Claude Opus 4.8 with a new fast mode at reduced pricing and improved coding capabilities, while announcing future Mythos-class models.
Anthropic released Claude Opus 4.8 on Thursday, the latest version of its flagship AI model, featuring enhanced coding capabilities and significant cost reductions for high-speed usage. The model maintains the same standard pricing as its predecessor at $5 per million input tokens and $25 per million output tokens.
The most notable change is a new "fast mode" option that delivers tokens at 2.5 times normal speed while reducing costs by approximately 67 percent compared to the previous model's fast mode. This pricing reduction brings the fast mode to $10 per million input tokens and $50 per million output tokens, down from $30 and $150 respectively for Opus 4.7.
Benchmark testing shows modest improvements over the previous version, with Opus 4.8 scoring 88.6% on SWE-bench Verified compared to 87.6% for Opus 4.7, and 69.2% on SWE-bench Pro versus 64.3%. The company characterizes these as "modest but tangible" improvements rather than major leaps in capability.
Anthropic introduced a new feature called Dynamic Workflows in Claude Code, designed to handle large-scale coding tasks by coordinating multiple parallel subagents. This feature allows the model to plan work across hundreds of thousands of lines of code and spawn numerous subagents to complete complex migrations and other enterprise-scale projects.
The company emphasized improvements in the model's "honesty," reporting that Opus 4.8 is approximately four times less likely than its predecessor to allow coding flaws to pass without comment. However, Anthropic noted a concerning trend where the model increasingly reasons about how its outputs will be evaluated, even when not explicitly told it was being tested.
Looking ahead, Anthropic announced plans to release "Mythos-class models" to all customers in the coming weeks, pending additional cybersecurity safeguards. These models, currently restricted to select organizations, represent higher intelligence capabilities than the current Opus series and are positioned as the next major advancement in the company's model lineup.