The upgrade arrives 41 days after Opus 4.7 as competition intensifies across the AI industry
Anthropic released Claude Opus 4.8 on May 28, 2026. The company described it as an upgraded model focused on software engineering, agent-based tasks, and long-form analytical work. It is available to all users through claude.ai, the Claude API, and Claude Code, with no change in pricing from the previous version.
Claude Opus 4.8: What’s New
Anthropic claims improvements across the following areas:
- Agentic coding
- Multidisciplinary reasoning
- Agentic computer use
- Knowledge work
- Agentic financial analysis
Benchmark comparisons with Opus 4.7 show agentic coding scores rising from 64.3% to 69.2%, and multidisciplinary reasoning with tools from 54.7% to 57.9%.
Claude Opus 4.8 Honesty and Reliability Improvements
This area received notable emphasis from Anthropic.
- Early testers reported Opus 4.8 is more likely to flag uncertainties and less likely to make unsupported claims.
- Anthropic’s evaluations show the model is around four times less likely than Opus 4.7 to let flaws in its own code pass without flagging them.
- The model scored well on what Anthropic called “prosocial” traits, including supporting user autonomy and acting in the user’s best interest, reaching a level similar to Mythos, Anthropic’s most advanced model.
- Bridgewater Associates cited the model’s tendency to proactively flag issues with inputs and outputs in early testing.
Claude Opus 4.8 New Features Explained
Dynamic Workflows
Claude Code now includes Dynamic Workflows, a feature that plans work, runs parallel sub-agents, verifies outputs, and reports back to the user. According to Anthropic, it can carry out codebase-scale migrations across hundreds of thousands of lines of code from kickoff to merge. This feature is available for Claude Code users on Enterprise, Team, and Max plans.
Effort Controls
Users can now adjust how much effort the model applies to a task. By default, Opus 4.8 runs on high effort, which Anthropic says uses similar tokens to Opus 4.7 on coding tasks while delivering better performance.
Fast Mode
A fast mode lets the model work at 2.5 times the speed. Standard pricing is $5 per million input tokens and $25 per million output tokens. Fast mode costs $10 per million input and $50 per million output tokens.
Claude Opus 4.8 Pricing
| Mode | Input (per million tokens) | Output (per million tokens) |
| Standard | $5 | $25 |
| Fast | $10 | $50 |
Pricing remains unchanged from Claude Opus 4.7.
Why Anthropic Released Claude Opus 4.8 So Quickly
The 41-day gap between Opus 4.7 and Opus 4.8 is unusually short for Anthropic. The fast turnaround may be connected to a lukewarm reception to Opus 4.7 among some users, and to fresh launches from OpenAI and Google during the same period.
What Comes After Claude Opus 4.8
Opus 4.8 still lags behind Mythos, Anthropic’s most advanced model, which is not yet publicly available. Anthropic said it expects to make Mythos-class models available to all customers in the coming weeks. The company also disclosed that its revenue run rate has more than tripled to $47 billion over the past three months.
Conclusion
Claude Opus 4.8 is a targeted upgrade with measurable gains in coding, agentic tasks, and model honesty. The addition of Dynamic Workflows and effort controls gives developers and enterprise customers more flexibility over cost and performance. With Mythos-class models reportedly weeks away from general availability, Opus 4.8 appears to be Anthropic’s immediate competitive answer while it prepares for a larger release.
