Unverified50% confidenceBenchmarkExact time

有独立评测机构测出Opus 5.5在60个模型的排名中位列第四，得分比上一代略低

1

Sources

50%

Confidence

Short-term (~14 days)

Relevance

9/30/2026

First Seen

Valid until: 10/14/2026

Sources

Anthropic与OpenAI同日降价：AI主战场从性能转向成本

Related Entities

Related Claims

Unverified在Artificial Analysis综合指数中，Opus 5.5得58分，Sol得48分，Luna得37分74% similar Unverified在MATH数据集的Level 5竞赛级难题上，前沿模型准确率仍低于60%70% similar Unverified同一个模型在不同评测平台上的排名可能大相径庭68% similar UnverifiedAnthropic称Opus 5.5比上一代快30%以上，默认设置下可省约40%算力67% similar Unverified数据显示'纯粹独立模态'出现的情况不足5%66% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/961368

API

curl https://kongchang.com/api/v1/knowledge/claims/961368

MCP

get_claim(id=961368)