Appendix A Appendix

本文为论文《Qwen3：兼顾思考与非思考模式的开源模型家族》「论文原文」板块的对应章节，按 arXiv 官方全文逐字保留，层级与「中文全文翻译」板块一一对应。

A.1 Additional Evaluation Results

A.1.1 Long-Context Ability

Table 23: Performance of Qwen3 Models on the RULER benchmark.

For evaluating long-context processing capabilities, we report the results on the RULER benchmark (Hsieh et al., 2024) in Table 23. To enable length extrapolation, we utilize YARN (Peng et al., 2023) with a scaling_factor=4. In thinking mode, we set the thinking budget to 8192 tokens to mitigate overly verbose reasoning on the extremely long inputs.

The results show that:

In non-thinking mode, Qwen3 outperforms Qwen2.5 models of a similar size in long-context processing tasks.

In thinking mode, the model’s performance slightly degrades. We hypothesize that the thinking content does not provide significant benefits for these retrieval tasks, which do not rely on reasoning and may instead interfere with the retrieval process. We are committed to enhancing the long-context capability in the thinking mode in future versions.

A.1.2 Multilingual Ability

Table 24-35 presents the detailed benchmark scores across various languages, including Spanish, French, Portuguese, Italian, Arabic, Japanese, Korean, Indonesian, Russian, Vietnamese, German, and Thai. The results of these tables demonstrate that the Qwen3 series models achieve competitive performance across all evaluated benchmarks, showcasing their strong multilingual capabilities.

To evaluate the performance of Qwen3 across a broader range of languages, we utilize Belebele (Bandarkar et al., 2023), a benchmark for natural language understanding. We conduct evaluations on 80 supported languages from the benchmark, excluding 42 unoptimized languages, as shown in Table 36 (organized by language family). The performance comparison between Qwen3 and other baseline models on the Belebele benchmark is presented in Table 37. The results show that Qwen3 achieves comparable performance to similarly-sized Gemma models while outperforming Qwen2.5 significantly.

Table 24: Benchmark scores for language: Spanish (es). The highest and second-best scores are shown in bold and underlined, respectively.

Table 25: Benchmark scores for language: French (fr). The highest and second-best scores are shown in bold and underlined, respectively.

Table 26: Benchmark scores for language: Portuguese (pt). The highest and second-best scores are shown in bold and underlined, respectively.

Table 27: Benchmark scores for language: Italian (it). The highest and second-best scores are shown in bold and underlined, respectively.

Table 28: Benchmark scores for language: Arabic (ar). The highest and second-best scores are shown in bold and underlined, respectively.

Table 29: Benchmark scores for language: Japanese (ja). The highest and second-best scores are shown in bold and underlined, respectively.

Table 30: Benchmark scores for language: Korean (ko). The highest and second-best scores are shown in bold and underlined, respectively.

Table 31: Benchmark scores for language: Indonesian (id). The highest and second-best scores are shown in bold and underlined, respectively.

Table 32: Benchmark scores for language: Russian (ru). The highest and second-best scores are shown in bold and underlined, respectively.

Table 33: Benchmark scores for language: Vietnamese (vi). The highest and second-best scores are shown in bold and underlined, respectively.

Table 34: Benchmark scores for language: German (de). The highest and second-best scores are shown in bold and underlined, respectively.

Table 35: Benchmark scores for language: Thai (th). The highest and second-best scores are shown in bold and underlined, respectively.

Table 36: Language families and language codes supported by Qwen3 in Belebele Benchmark

Table 37: Comparison of Belebele Benchmark performance between Qwen3 and other baseline models. Scores are highlighted with the highest in bold and the second-best underlined.