Claude Opus 4.8 launches with major upgrades, Fast Mode

Anthropic launched Claude Opus 4.8, featuring major coding and reasoning upgrades, Fast Mode for lower costs, and improved user control. Claude Mythos, its cybersecurity AI, is nearing wider release after additional testing and safeguards.

Anthropic has unveiled Claude Opus 4.8, marking a notable upgrade over Opus 4.7. The latest model delivers enhanced coding and reasoning capabilities, outperforming OpenAI’s GPT-5.5 and Google’s Gemini 3.1 Pro in several benchmark tests. Key improvements include a fourfold reduction in undetected coding errors, advanced agentic abilities, and a new effort control feature that lets users adjust processing depth according to task needs. The integrated Fast Mode reportedly slashes user costs by a factor of three compared to previous standard tiers, though the baseline for this comparison remains unclear. Pricing for standard use is unchanged at $5 per million input tokens and $25 per million output tokens. Additionally, Anthropic announced that its cybersecurity-focused Claude Mythos AI model is approaching a broader release. After further testing and the addition of new safeguards—addressing concerns about its autonomous vulnerability identification and cyberattack execution—wider access to Mythos is expected in the coming weeks.