Anthropic · Proprietary · tested 8 Nov 2026
Claude Opus 5.5 High effortCVP Defense Access
Tested with verified cyber access: CVP Defense Access, the Defense Access tier of Anthropic's Cyber Verification Program. It's the same Claude Opus 5.5 as the public API, with Anthropic's cyber safeguards relaxed for verified security work. Charts mark it with a checkered fill.
Safety refusals affected this result
Anthropic's cyber safeguards stopped attempts in 32 of 100 labs, refusing or answering with Claude Opus 4.8 instead. Stopped attempts score zero. BB Arena never counts a fallback model's work: the result is this model's own. BB Arena is enrolled in Anthropic's cyber program for verified security researchers, and these refusals happened anyway.
With refusals in more than 10 of 100 labs, its score says more about the safeguards than the model, so it isn't ranked: it's on the refusals board. Its cost, tokens and time are left out of comparisons too. How refusals are scored
On the labs it was never refused
Anthropic's safeguards stopped attempts in 32 of 100 labs. On the other 68, Claude Opus 5.5 worked without refusals, so here it's compared like for like with the ranked models that did best on those same 68 labs. These labs aren't a random sample of the benchmark, so this doesn't predict a full score.
- Claude Opus 5.5 (High)not ranked7 of 68 solved89 of 272 rungs
- Grok 4.6 (High)#2 overall17 of 68 solved132 of 272 rungs
- Claude Opus 5 (High)#1 overall16 of 68 solved123 of 272 rungs
- Grok 4.7 (High)#4 overall16 of 68 solved110 of 272 rungs
- DeepSeek V4.1 Flash (Max)#3 overall13 of 68 solved114 of 272 rungs
By vulnerability class
How Claude Opus 5.5 (High) did on each type of vulnerability. Shown as counts: each class has 10 labs, so one lab is 10% of a class.
- Access controlSafety refusals in 2 of 10 labs1 of 10 solved15 of 40 rungs
- Authentication & sessionsSafety refusals in 2 of 10 labs0 of 10 solved12 of 40 rungs
- Injection: queries & dataSafety refusals in 4 of 10 labs2 of 10 solved12 of 40 rungs
- Code executionSafety refusals in 3 of 10 labs0 of 10 solved5 of 40 rungs
- Server-side accessSafety refusals in 4 of 10 labs0 of 10 solved7 of 40 rungs
- Browser-sideSafety refusals in 7 of 10 labs0 of 10 solved4 of 40 rungs
- Infrastructure & edge0 of 10 solved11 of 40 rungs
- Business logic & racesSafety refusals in 1 of 10 labs1 of 10 solved11 of 40 rungs
- Secrets, crypto & supply chainSafety refusals in 5 of 10 labs3 of 10 solved17 of 40 rungs
- AI agentsSafety refusals in 4 of 10 labs0 of 10 solved6 of 40 rungs
Provider performance
Measured per single agent against the provider endpoint used for the run.
- Response time
- 6.62 s median · 19.52 s p95
- Output speed
- 78.3 tok/s
- API success rate
- 99.88%
- Input context
- 19.7K median · 54.7K p95
About this result
Every model runs in the same neutral harness, so results compare the models themselves.
- Harness
- BB Arena neutral harness
- Safety refusals
- 32 of 100 labs