Qwen 3.8 users report better results forcing reasoning effort to low, llama.cpp flag shared
Jorlen · reddit · 2026-09-03
Reddit users discuss whether Qwen 3.8's (flash next and 27b) default extra-high reasoning effort is counterproductive, with many reporting better results forcing it to low. The OP shares the llama.cpp flag --chat-template-kwargs '{"reasoningeffort":"low"}' and asks how others configure it.
More from coding & agent
- LukeW built an agent-orchestration app for a foldable: managing agents is now an all-day job — LukeW · 2026-10-21
- Agentic SDLC: engineers shift from implementers to agent orchestrators as cycles shrink from weeks to hours — Pavan_Belagatti · 2026-10-21
- OpenAI's dots can now use company tools via Merge for Workforce with employee-bound permissions — shensi · 2026-10-21
- The topology of an LLM agent: where the model, loop, and tools run shapes cost and security — davidcrawshaw · 2026-10-21
- Agents with full-disk and accessibility access are an enterprise security nightmare, argues Runlayer — leveragedupside · 2026-10-21
- Acellera runs an unsupervised virtual lab of AI scientists on a real IL-23 molecular discovery program — gdefabritiis · 2026-10-21