Qwen/Qwen3-4B-Instruct-2507

Qwen3-4B-Instruct-2507: Enhanced Instruction-Following LLM

Qwen3-4B-Instruct-2507 is an updated 4.0 billion parameter instruction-tuned causal language model from Qwen, building upon the Qwen3-4B non-thinking mode. It features a substantial native context length of 262,144 tokens, enabling advanced long-context understanding.

Key Capabilities and Enhancements

• General Capabilities: Significant improvements in instruction following, logical reasoning, text comprehension, mathematics, science, coding, and tool usage.
• Knowledge & Multilingualism: Substantial gains in long-tail knowledge coverage across multiple languages, as evidenced by strong performance on MMLU-ProX and PolyMATH benchmarks.
• Alignment & Subjectivity: Markedly better alignment with user preferences in subjective and open-ended tasks, leading to more helpful responses and higher-quality text generation.
• Agentic Use: Excels in tool-calling capabilities, with recommendations to use Qwen-Agent for optimal integration.

Performance Highlights

The model demonstrates strong performance across various benchmarks, often surpassing its predecessor, Qwen3-4B Non-Thinking, and in some cases, even larger models. Notable improvements include:

• Knowledge: Achieves 69.6 on MMLU-Pro and 84.2 on MMLU-Redux.
• Reasoning: Scores 47.4 on AIME25 and 80.2 on ZebraLogic.
• Coding: Reaches 35.1 on LiveCodeBench v6 and 76.8 on MultiPL-E.
• Alignment: Scores 83.4 on IFEval and 83.5 on Creative Writing v3.

Recommended Use Cases

This model is particularly well-suited for applications requiring:

• Advanced instruction following and complex reasoning.
• Long-context understanding and generation.
• High-quality, aligned responses in subjective and open-ended conversational scenarios.
• Agentic workflows and tool-use integration.