How are you serving them? This is most often caused by an incorrect template. (the thing that tells the inference engine how to parse the think/tool parts of th

How are you serving them? This is most often caused by an incorrect template. (the thing that tells the inference engine how to parse the think/tool parts of the answer) For vllm the official recipe [1] (for another model in the same family) is this: --reasoning-parser qwen3 \ --enable-auto-tool-choice \ --tool-call-parser lfm2 [1] - https://recipes.vllm.ai/LiquidAI/LFM2.5-8B-A1B https://recipes.vllm.ai/LiquidAI/LFM2.5-8B-A1B