qwen3-30b-a3b-fp8

qwen3-30b-a3b-fp8 is Qwen’s open-weight mixture-of-experts model designed for efficient reasoning, coding, multilingual text generation, and agentic workflows. It activates only 3B of its 30B parameters per token, delivering strong model quality with lower latency and inference cost. It supports function calling, streaming, and multilingual workloads, making it well suited for agents, code assistance, automation, structured data tasks, and high-volume production applications. On ZeroGPU, it runs across the hybrid inference cloud to optimize speed and cost for production workloads.