Claude Opus 4.5
FrontierAnthropic•Released on 2025-12-01
Anthropic's flagship model, widely recognized as the top coding model. Excels at complex refactoring, large codebase comprehension, and agentic coding. Claude Code makes it the go-to choice for professional developers.
90
Overall Score
Voice of the community
sample 1,456“Claude 4 Opus is the most tasteful coder among all the frontier models. Leads in raw performance and prompt adherence.”
“Debugging is scary good - it literally found a memory leak in my React app that I'd been hunting for days.”
“Opus 4.5 is so good that everything else is well… noise. Once you get a hang of Skill creation and use, nothing else comes close.”
Core Specs
200K
Context Window
64K
Max Output
ReasoningOpen Sourcetextimage
Scenario Scores
Pros & Cons
Sentiment82% +12% ·6% −
Pros
- +Top-tier coding ability
- +Highest code quality
- +Strong context understanding
- +Mature Claude Code ecosystem
Cons
- −Highest pricing ($15/1M input)
- −Average response speed
- −Strict quota limits
Pricing
Input (per 1M tokens)$5.00
Output (per 1M tokens)$25.00
Subscription$20/month
Updated on 2026-03-06
Get Started with Claude
💡 Free tier uses Sonnet. Upgrade to Pro for Opus.
Benchmarks
sweBench73.3%
mmlu92.1%
humanEval96.5%
Reliability
SLA99.9%
Incidents (30d)3
Last Incident2026-05-04
与 Opus 4.6 共享基础设施；5/4 08:09-09:49 UTC (~1h40m) Opus 4.5 专属 elevated errors，已 resolved；5/4 13:59-14:45 UTC (~46min) Opus 4.5 + Sonnet 4.5 同时 elevated errors，已 resolved (status.claude.com)
View Status Page →