Unverified50% confidenceFactExact time
Thinking mode is enabled by default at high intensity; only three levels actually take effect: minimal maps to low, medium and xhigh map to high, and ultra maps to max.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/19/2026
First Seen
Valid until: 1/17/2027
Sources
Related Entities
Related Claims
Unverified推理强度档位（low/medium/high/xhigh）本质上是控制模型被允许生成的思维链token数上限，档位越高等待时间越长72% similarUnverifiedThe reasoning tiers (low, medium, high, X high) set an upper token limit for thinking—the model may use fewer tokens but cannot exceed the limit, rather than enforcing a fixed amount of thinking.70% similarVerifiedMax mode raises the lower bound of reasoning tokens rather than the upper bound, forcing the model to reach a certain amount of reasoning before it is considered complete, which often harms performance.65% similarVerifiedThe Reasoning Effort parameter is divided into levels such as low, medium, high, and extra high; the higher the level, the stronger the reasoning capability, but token consumption and latency also increase proportionally.62% similarUnverifiedreasoningEffort generally supports three levels — 'low', 'medium', and 'high' — corresponding to different depths of reasoning and token consumption.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1069321API
curl https://kongchang.com/api/v1/knowledge/claims/1069321MCP
get_claim(id=1069321)