Model Overview
Description:
The RadixArk Qwen3.8-27B-NVFP4 model is the quantized version of
Qwen/Qwen3.8-27B. The quantization was produced at RadixArk using
NVIDIA Model Optimizer, following a mixed NVFP4 W4A4 recipe.
Run on SGLang: launch command and per-platform recipes in the
Qwen3.8-27B cookbook.
Third-Party Community Consideration
This model is not owned or developed by RadixArk. It is a quantized derivative of Qwen's model; see the upstream
Qwen3.8-27B model card for the source model's capabilities, training information, limitations, and license.
License/Terms of Use:
Deployment Geography:
Global
Use Case:
Developers looking to deploy an off-the-shelf, pre-quantized model in AI agent systems, chatbots, RAG systems, and other AI-powered applications.
Release Date:
Model Architecture:
Architecture Type: Transformers (Dense Multimodal)
Network Architecture: Qwen3.8-27B
Number of Model Parameters: 27B
Input:
Input Type(s): Text, image, and video
Input Format(s): String and visual media
Other Properties Related to Input: Native context length up to 262,144 tokens.
Output:
Output Type(s): Text
Output Format: String
Software Integration:
Supported Runtime Engine(s):
Supported Hardware Microarchitecture Compatibility:
- NVIDIA Blackwell (this checkpoint was produced and validated on GB300)
Preferred Operating System(s):
Model Version(s):
Quantized with
NVIDIA Model Optimizer, commit
87c9f8cf83021957d1a1a575c90c9a4eaaf7ef0c.
Training, Testing, and Evaluation Datasets:
Calibration Data:
Calibration used 1,024 samples from the abisee/cnn_dailymail training split with sequence length 512.
Training Dataset:
RadixArk did not train or fine-tune this checkpoint. Training information is inherited from the upstream
Qwen3.8-27B model card.
Evaluation Dataset:
The model was evaluated on GSM8K and Terminal-Bench 2.1.
Post Training Quantization
The MLP gate_proj, up_proj, and down_proj layers and lm_head use dynamic NVFP4 W4A4 quantization with group size 16. Attention weights use FP8, while MTP and vision tensors retain the source BF16 precision.
Usage
The following SGLang configuration uses four NVIDIA Blackwell GPUs:
1sglang serve \
2 --trust-remote-code \
3 --model-path RadixArk/Qwen3.8-27B-NVFP4 \
4 --tp-size 4 \
5 --mem-fraction-static 0.75 \
6 --speculative-algorithm NEXTN \
7 --speculative-num-steps 3 \
8 --speculative-eagle-topk 1 \
9 --speculative-num-draft-tokens 4 \
10 --reasoning-parser qwen3 \
11 --tool-call-parser qwen3_coder \
12 --host 0.0.0.0 \
13 --port 30000
For other deployment topologies and hardware-specific configurations, see the
SGLang Qwen3.8-27B cookbook.
Evaluation
The benchmark results below were produced with this NVFP4 checkpoint on 4x NVIDIA B300/GB300 GPUs using TP4 SGLang deployments.
| Benchmark | Evaluation protocol | Score |
|---|
| GSM8K | Full 1,319-example split, thinking mode, sgl-eval | 97.27% (1,283/1,319) |
| Terminal-Bench 2.1 | 84-task subset, Claude Code 2.1.228, pass@1 | 73.81% (62/84) |
GSM8K used temperature=1.0, top_p=0.95, and top_k=20. The reported evaluations were text-only.
Model Limitations:
The base model may generate inaccurate, incomplete, irrelevant, biased, or otherwise undesirable responses. Developers should evaluate the model for their intended use case and apply appropriate safeguards.
Ethical Considerations
RadixArk believes trustworthy AI is a shared responsibility. Developers should ensure that use of this model complies with the upstream license and meets the safety, privacy, and reliability requirements of their application.
