Claude Opus 4.8 arrives with effort controls and dynamic workflows
Anthropic's new Opus 4.8 model adds a user-set effort control, Claude Code dynamic workflows, and a faster, cheaper fast mode at the same price as before.
What Anthropic announced
Anthropic is upgrading Claude Opus to a new version, Claude Opus 4.8. The company says it builds on Opus 4.7 with improvements across benchmarks and is a more effective collaborator, available today for the same price. The release lands alongside several new features: users on claude.ai now have control over the amount of effort Claude puts into a task, Claude Code has a new dynamic workflows feature for very large-scale problems, and fast mode for Opus 4.8, where the model can work at 2.5 times the speed, is now three times cheaper than it was for previous models.
Faster, cheaper, and more capable
Anthropic defaults Opus 4.8 to high effort, which it calls the best overall balance of quality and user experience; on coding tasks this spends a similar number of tokens as Opus 4.7's default but with better performance. Users can also choose "extra" (called "xhigh" in Claude Code) or "max" effort when they want the model to spend more tokens. Dynamic workflows, available in research preview for Claude Code Enterprise, Team, and Max plans, let Claude plan work and then run hundreds of parallel subagents in a single session before verifying its own outputs. Anthropic says Claude Code with Opus 4.8 can now carry out codebase-scale migrations across hundreds of thousands of lines of code from kickoff to merge, using the existing test suite as its bar.
Benchmark and customer results
Anthropic's post shares quotes from early testers about their own results rather than a single official scorecard. One tester said that on their own Super-Agent benchmark, Opus 4.8 was the only model to complete every case end-to-end, beating prior Opus models and GPT-5.5 at parity on cost. Another, using CursorBench, said it exceeds prior Opus models across every effort level. A tester in legal work said it delivers the highest score recorded on their Legal Agent Benchmark and is the first model to break 10% overall on the all-pass standard. Another tester called it the strongest computer-use and browser-agent model they had tested, citing a score of 84% on Online-Mind2Web, a jump over both Opus 4.7 and GPT-5.5. Databricks reports that in its Genie agent, the new model lets it reason directly over PDFs, diagrams, and other unstructured content at 61% cheaper token cost than Opus 4.7, and a tester working on Devin says the release fixes comment-verbosity and tool-calling issues seen with Opus 4.7.
Honesty and alignment
Anthropic says one of the most prominent improvements in Opus 4.8 is its honesty. Early testers report that Opus 4.8 is more likely to flag uncertainties about its work and less likely to make unsupported claims, and Anthropic's evaluations show it is around four times less likely than its predecessor to let flaws in code it has written pass unremarked. Anthropic's Alignment team also concluded that Opus 4.8 reaches new highs on measures of prosocial traits such as supporting user autonomy, with rates of misaligned behavior substantially lower than Opus 4.7 and similar to Anthropic's best-aligned model, Claude Mythos Preview.
Key facts
- Claude Opus 4.8 builds on Opus 4.7 and is available today at the same price.
- Fast mode now runs at 2.5 times the speed and is three times cheaper than for previous models.
- Opus 4.8 scored 84% on Online-Mind2Web and broke 10% overall on the Legal Agent Benchmark's all-pass standard.
- Dynamic workflows, in research preview for Claude Code Enterprise, Team, and Max plans, can run hundreds of parallel subagents in one session.
- The Messages API now accepts system entries inside the messages array, so developers can update instructions mid-task without breaking the prompt cache.
What developers get
Beyond the model itself, Anthropic is shipping an effort control for claude.ai and Cowork on all plans, letting users trade faster responses and slower rate-limit use against deeper thinking. The Messages API change lets a harness update permissions, token budgets, or environment context as an agent runs, without routing the update through a user turn or breaking the prompt cache.
Check your understanding
What two new features does Anthropic say ship alongside the Opus 4.8 model itself?
According to Anthropic, how much faster and how much cheaper is the new fast mode compared with previous models?