Skip to content
Anthropic

Claude Opus 4.8

Frontier

AnthropicReleased on 2026-05-28

Anthropic's most capable generally available model, released May 28, 2026 — weeks after Opus 4.7. Anthropic describes it as having sharper judgement, more honesty about its own progress, and the ability to work independently for longer. It is roughly 4x less likely than Opus 4.7 to miss flaws in code it produces and less prone to unsupported claims. Same pricing as Opus 4.7 makes it a drop-in upgrade, and a new Fast Mode runs the same model at higher speed.

92
Overall Score

Voice of the community

sample 40

Claude Opus 4.8 has sharper judgement, more honesty about its progress, and the ability to work independently for longer than its predecessors.

anthropic.com/news2026-05-28

Opus 4.8 is roughly four times less likely than Opus 4.7 to miss flaws in code it produces, and Anthropic calls it its 'most honest' model yet.

Inc.2026-05-28

Opus 4.8 beats Opus 4.7 and GPT-5.5 on many benchmarks, leading agentic coding on SWE-Bench Pro at 69.2%.

OfficeChai2026-05-28

Core Specs

1000K
Context Window
128K
Max Output
ReasoningOpen Sourcetextimage

Pros & Cons

Sentiment75% +20% ·5% −

Pros

  • +Leads agentic coding (SWE-Bench Pro 69.2 vs 4.7's 64.3, GPT-5.5 58.6)
  • +~4x less likely to miss flaws in its own code
  • +Best-in-class multidisciplinary reasoning (Humanity's Last Exam)
  • +Same price as Opus 4.7 (drop-in upgrade)
  • +1M context window + new Fast Mode for higher speed

Cons

  • Still 2x the price of comparable GPT-5.5 tiers
  • Launch benchmarks are Anthropic-reported, pending independent verification
  • Fast Mode costs 2x the standard per-token price

Pricing

Input (per 1M tokens)$5.00
Output (per 1M tokens)$25.00
Subscription$20/month
Updated on 2026-05-29

Get Started

1Visit the provider's website
2Create an account
3Start using the model

Benchmarks

noteAnthropic-reported launch benchmarks (pending independent verification). Opus 4.8 leads agentic coding and reasoning vs Opus 4.7, GPT-5.5 and Gemini 3.1 Pro.%
swe-bench-pro69.2%
humanitys-last-exam-no-tools49.8%
humanitys-last-exam-with-tools57.9%
osworld-verified83.4%

Reliability

SLA99.9%
Incidents (30d)0
View Status Page →