Qwen3.8 2.4T A95B

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Alibaba with 95 billion active parameters out of 2.4 trillion total parameters.

Strengths

The sparse architecture activates only the necessary parameters per token, reducing compute requirements compared to dense models of equivalent capacity.

Best for

Workloads requiring substantial reasoning capacity where inference efficiency matters, particularly when deployed on infrastructure that supports sparse computation.

Limitations

Open-weight sparse models may require specialized hardware or software to realize efficiency gains; the model sits between Qwen3.6 35B A3B and Qwen3.8 Max in capability, lacking the multimodal features or optimizations of the flagship Max variant.