Extended thinking off; 200K context and default sampling are stated for MCP Atlas; Table 2.3.A applies its no-thinking footnote to MCP Atlas, τ² Retail and τ² Telecom.
未知项
The system card does not state an independent effort level for these no-thinking evaluations.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
τ²-bench Retail。
条件
No extended thinking per Table 2.3.A footnote 4; 5-trial average per Table 2.3.A. Table 2.8.A also identifies the Opus 4.1 user simulator and task prompt addendum.
Differences are calculated only for unique, active, exact values between direct predecessors when the unit, metric, and explicitly recorded configuration and conditions match.
· 1 score record for this release, including 1 exact value (may include historical records)
Claude Opus 4.5 · τ²-bench scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Success rate88.9%Reported 88.9% · Unit %
Published Unconfirmed
No extended thinking
Settings
Extended thinking off; 200K context and default sampling are stated for MCP Atlas; Table 2.3.A applies its no-thinking footnote to MCP Atlas, τ² Retail and τ² Telecom.
Unknowns
The system card does not state an independent effort level for these no-thinking evaluations.
Tools
Tool access not recorded
Reasoning budget
not recorded
Framework
not recorded
Protocol
τ²-bench Retail。
Conditions
No extended thinking per Table 2.3.A footnote 4; 5-trial average per Table 2.3.A. Table 2.8.A also identifies the Opus 4.1 user simulator and task prompt addendum.