Model techniques map
Techniquesinference & servingreasoning control

implementation detail · filed under inference & serving

Reasoning-effort resolution and system-prompt injection

A chat template resolves the requested effort to a supported level and inserts that level into the system prompt.

source
1
model
1
lab adopt it
1
strongest
default

How sources treat it

One count per evidence span, weakest treatment to strongest.

default 1

Documented in

Evidence

1 span quoted from the sources, strongest treatment first.

The chat template resolves effort to max unless reasoning_effort is explicitly "low" or "high" (any other value falls back to max), then injects Reasoning Effort: Low|High|Max into the system prompt.

defaultinference servingin GLM-5.3-FlashZ.ai

Filed alongside

Other methods under inference & serving :: reasoning control.