GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
mkd-hossain
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
mkd-hossain
Model Tree
mkd-hossain/Keural-Nova-v1.2-experimental
Input Modalities
Output Modalities
Supported Functionality
FP4 tensor cores exist only on Blackwell (B200 / GB200 / RTX 50xx). Serve there for the speed/memory benefit. (Hopper/H200 lack FP4 kernels — this checkpoint is produced on H200 but runs accelerated on Blackwell.)
bash
vllm serve mkd-hossain/Keural-Nova-v1.2-experimental-NVFP4 \--served-model-name Keural-Nova-v1.2 \--tensor-parallel-size 2 \--max-model-len 262144 \--tool-call-parser qwen3_xml --enable-auto-tool-choice
llm-compressor (compressed-tensors), NVFP4 scheme.lm_head + MTP excluded (bf16).NVFP4 typically costs ~1–2% vs bf16 — the same small delta as the public
nvidia/RedHatAI NVFP4 Qwen3.6-A3B checkpoints. Validate on your Blackwell against the
bf16 base (Keural-Nova-v1.2-experimental)
on your workload (tool-calling, Korean, code) before production.
Apache-2.0, inheriting the base Qwen/Qwen3.6-35B-A3B license. Quantization recipe after
RedHatAI/Qwen3.6-35B-A3B-NVFP4. "Keural Nova" is a model by MKD.