Anthropic has released Claude Opus 4.8, its latest flagship model, just 41 days after Opus 4.7 — an unusually fast upgrade cycle likely driven by mixed reception to the previous version and competitive pressure from OpenAI and Google. The new model emphasizes better handling of uncertain or bad data, with early testers noting it proactively flags issues in inputs and outputs. Alongside the model, Anthropic launched Dynamic Workflows in research preview, a feature enabling Opus to coordinate hundreds of parallel subagents for large-scale tasks like codebase-wide migrations via Claude Code. The company also hinted that its more advanced Mythos model, held back due to cybersecurity concerns, may be released to all customers within weeks.
Sort: