Model Detail

Qwen3.6-27B-PrismaSCOUT-Blackwell-NVFP4-BF16-vllm is an AI model with 27B parameters released by rdtand. Distributed under the permissive apache-2.0 license.

Qwen3.6-27B-PrismaSCOUT-Blackwell-NVFP4-BF16-vllm ships with 27B parameters. Total weight footprint is approximately 17.0 GB, which is the relevant figure when planning local-inference VRAM. The apache-2.0 license is permissive, allowing commercial deployment and derivative work without per-seat fees, though attribution requirements still apply.

Qwen3.6-27B-PrismaSCOUT-Blackwell-NVFP4-BF16-vllm is best fit for general-purpose AI workloads. Treat this as a starting matrix rather than a benchmark verdict — the right deployment usually depends on the specific evaluation suite that mirrors your workload.