Nathan(@nathanhabib1011):🚨 ITS HERE🚨 ⚡Qwen3.8-27B - Multimodal dense model. With just 27B parameters, it outperforms most models on the @huggingface leaderboards.. - Built for builders: Apache 2.0 license Swe-Bench Pro, Deep-Swe and Claw-Eval, those 3 benchmarks evaluate how good a model is at being an agent, needless to say @Alibaba

2026.08.14 15:23

🚨 ITS HERE🚨 ⚡Qwen3.8-27B - Multimodal dense model. With just 27B parameters, it outperforms most models on the @huggingface leaderboards.. - Built for builders: Apache 2.0 license Swe-Bench Pro, Deep-Swe and Claw-Eval, those 3 benchmarks evaluate how good a model is at being an agent, needless to say @Alibaba_Qwen team delivered..