Qwen3 VL 4B Instruct

Model Overview

Qwen3-VL is the most powerful vision-language model in the Qwen series to date.

Key Enhancements:

• Visual Agent: Operates PC/mobile GUIs—recognizes elements, understands functions, invokes tools, completes tasks.

• Visual Coding Boost: Generates Draw.io/HTML/CSS/JS from images/videos.

• Advanced Spatial Perception: Judges object positions, viewpoints, and occlusions; provides stronger 2D grounding and enables 3D grounding for spatial reasoning and embodied AI.

• Long Context & Video Understanding: Native 256K context, expandable to 1M; handles books and hours-long video with full recall and second-level indexing.

• Enhanced Multimodal Reasoning: Excels in STEM/Math—causal analysis and logical, evidence-based answers.

• Upgraded Visual Recognition: Broader, higher-quality pretraining is able to “recognize everything”—celebrities, anime, products, landmarks, flora/fauna, etc.

• Expanded OCR: Supports 32 languages (up from 19); robust in low light, blur, and tilt; better with rare/ancient characters and jargon; improved long-document structure parsing.

• Text Understanding on par with pure LLMs: Seamless text–vision fusion for lossless, unified comprehension.

Model Architecture:

• Interleaved-MRoPE: Full‑frequency allocation over time, width, and height via robust positional embeddings, enhancing long‑horizon video reasoning.

• DeepStack: Fuses multi‑level ViT features to capture fine‑grained details and sharpen image–text alignment.

• Text–Timestamp Alignment: Moves beyond T‑RoPE to precise, timestamp‑grounded event localization for stronger video temporal modeling.

Model Source: Qwen/Qwen3-VL-4B-Instruct

License: apache-2.0

Multi Model QPC Configuration # 1

Precision	SoCs / Tensor slicing	NSP-Cores (per SoC)	Full Batch Size	Chunking Prompt Length	Context Length (CL)	CCL_Enabled	QPC URL	QPC Size	QPC Download	Onnx URL	Onnx Download	Generation Date
MXFP6	1	8	1	512	4096	False	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Encoder_qpc_8cores_1bs_512pl_4096cl_1devices_sdk1_21_6.tar.gz	796MB	Download	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Encoder_ONNX_sdk1_21_6.tar.gz	Download	23-July-2026
MXFP6	8	8	1	512	4096	False	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Decoder_qpc_8cores_1bs_512pl_4096cl_8devices_sdk1_21_6.tar.gz	16GB	Download	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Decoder_ONNX_sdk1_21_6.tar.gz	Download	23-July-2026

Multi Model QPC Configuration # 2

Precision	SoCs / Tensor slicing	NSP-Cores (per SoC)	Full Batch Size	Chunking Prompt Length	Context Length (CL)	CCL_Enabled	QPC URL	QPC Size	QPC Download	Onnx URL	Onnx Download	Generation Date
MXFP6	1	8	1	512	4096	False	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Encoder_qpc_8cores_1bs_512pl_4096cl_1devices_sdk1_21_6.tar.gz	796MB	Download	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Encoder_ONNX_sdk1_21_6.tar.gz	Download	24-July-2026
MXFP6	2	8	1	512	4096	False	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Decoder_qpc_8cores_1bs_512pl_4096cl_2devices_4instances_sdk1_21_6.tar.gz	6.2GB	Download	https://dc00tk1pxen80.cloudfront.net/SDK1.21.6/Qwen/Qwen3-VL-4B-Instruct/Qwen_Qwen3-VL-4B-Instruct_Decoder_ONNX_sdk1_21_6.tar.gz	Download	24-July-2026

Run This Model

Download QPCs

mkdir -p Qwen/Qwen3-VL-4B-Instruct
cd Qwen/Qwen3-VL-4B-Instruct

# Download Encoder QPC
wget <Encoder_QPC_Download_URL>
tar xzvf <encoder_qpc_filename.tar.gz>

# Download Decoder QPC
wget <Decoder_QPC_Download_URL>
tar xzvf <decoder_qpc_filename.tar.gz>

Run QPC

Replace <encoder_qpc_path> and <decoder_qpc_path> with the actual extracted QPC directories.

  python3 -m qaic_disagg  \
  --encode-port 4912  \
  --encode-device-group 0:1  \
  --decode-port 4916  \
  --decode-device-group 8:16  \
  --port 4921  \
  --model "Qwen/Qwen3-VL-4B-Instruct"  \
  --prefill-max-seq-len-to-capture 512  \
  --decode-max-seq-len-to-capture 512  \
  --max-model-len 4096  \
  --encode-override-qaic-config "height=364 width=532 num_cores=8 num_frames=5 kv_offload=True qpc_path=<encoder_qpc_path>"  \
  --decode-override-qaic-config "height=364 width=532 num_cores=8 num_frames=5 kv_offload=True qpc_path=<decoder_qpc_path>" \
  --encode-max-num-seqs 1  \
  --decode-max-num-seqs 1  \
  --quantization mxfp6  \
  --kv-cache-dtype mxint8  \
  --proxy_worker 3  \
  --limit-mm-per-prompt '{"image":5}'  \
  --enable-prefix-caching  \
  --show-hidden-metrics-for-version=0.7  \
  --router-policy least_outstanding   \
  --build-grafana-json  

Useful Links

• Environment Setup