Qwen3 introduces the updated version of the Qwen3-4B non-thinking mode, named Qwen3-4B-Instruct-2507, featuring the following key enhancements:

• Significant improvements in general capabilities, including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage.
• Substantial gains in long-tail knowledge coverage across multiple languages.
• Markedly better alignment with user preferences in subjective and open-ended tasks, enabling more helpful responses and higher-quality text generation.
• Enhanced capabilities in 256K long-context understanding.

Multi-size Context Performance Table

The model supports multiple context window sizes and automatically scales to the most appropriate one based on the input length.

Supported chips and their specific performance are listed in the table below:

Model Farm provides optimized model resources and test code, which can be obtained through the following two methods:

• Obtain via Model Farm page: Click Models & Test Code in the Performance Reference section on the right to obtain model resources and code packages.

• Obtain via command line (Recommand): Users with APLUX development boards can obtain model resources and code packages through the built-in MMS tool.

# Search Models
mms list [model name]

# Get Models
mms get -m [model name] -p [precision] -c [soc] -b [backend] -d [file path]

For MMS usage, please refer to: MMS Usage & Access to Preview Models

Qwen3-4B-Instruct-2507 has the following features:

• Type: Causal Language Models
• Training Stage: Pretraining & Post-training
• Number of Parameters: 4.0B
• Number of Paramaters (Non-Embedding): 3.6B
• Number of Layers: 36
• Number of Attention Heads (GQA): 32 for Q and 8 for KV
• Context Length: 262,144 natively.

NOTE: This model supports only non-thinking mode and does not generate <think></think> blocks in its output. Meanwhile, specifying enable_thinking=False is no longer required.

Note: This table showed source model instead of quantized model evaluation. Source Model Evaluation refer to Qwen3-4B-Instruct-2507 Evaluation Result

*: For reproducibility, we report the win rates evaluated by GPT-4.1.

Please refer to the model inference tags at the top to select an appropriate inference SDK.

Recommended to use AidGen & AidGenSE for inference