Model techniques map
Techniquesinference & servingreasoning control

specific method · filed under inference & serving

Interleaved thinking between tool calls

Reasoning is interleaved with tool calls, with evidence also noting that thinking can be enabled or disabled per request.

Also called Interleaved thinking, per-request control via enable_thinking.

sources
3
models
2
labs adopt it
2
strongest
core

How sources treat it

One count per evidence span, weakest treatment to strongest.

optional 2default 1core 1

Documented in

Evidence

4 spans quoted from the sources, strongest treatment first.

Native reasoning support: interleaved thinking between tool calls

coreotherin Laguna S 2.1Poolside

By default, only the thinking blocks generated in handling the latest user message is retained, resulting in a pattern commonly as interleaved thinking.

defaultinference servingin Qwen3.6-35B-A3BQwen

Native reasoning support: Interleaved thinking between tool calls, with the ability to enable or disable thinking per request.

optionalinference servingin Laguna XS 2.1Poolside

per-request control via enable_thinking

optionalotherin Laguna S 2.1Poolside

Filed alongside

Other methods under inference & serving :: reasoning control.