Anthropic ships Claude Opus 4.8 with parallel-subagent workflows and a 2.5x fast mode
Opus 4.8 introduces dynamic workflows — in research preview — that coordinate hundreds of parallel subagents in a single Claude Code session, alongside effort control and a fast mode at roughly 2.5x speed; benchmark results include 84% on Online-Mind2Web and the first pass at 10% on the Legal Agent Benchmark all-pass standard, a threshold no prior model had cleared. Pricing holds at $5/$25 per million tokens, with fast mode at $10/$50 — large-scale parallel orchestration lands inside the standard model tier, not a separate product.
Source: anthropic.com ↗
Opus 4.8 is the first model to break 10% overall on the all-pass standard. For substantive legal work, that's the kind of accuracy lift that translates directly into how much real attorney work our customers can hand off with confidence.
Why this matters
- → Opus 4.8 is the first model to break 10% on Legal Agent Benchmark all-pass standard, enabling lawyers to delegate substantive work with higher confidence.
- → Dynamic workflows let Claude orchestrate hundreds of parallel subagents in a single session, unlocking codebase-scale migrations previously infeasible.
- → Fast mode now costs 3x less than prior versions while delivering 2.5x speed, making large-scale automation economically viable at standard pricing tiers.