Qwen: Qwen3.6 Flash

Qwen3.6 Flash, released by qwen on April 26, 2026, is a fast and efficient multimodal language model with a substantial 1M token context window. It consistently ranks among the fastest models, achieving an Infinityth percentile across 8 benchmarks, and offers highly competitive pricing, also at an Infinityth percentile across 4 benchmarks. The model demonstrates strong reliability with an 88% success rate across 8 benchmarks, indicating consistent operational stability. In terms of benchmark performance, Qwen3.6 Flash shows a notable strength in "Hallucinations (Baseline)" with 96.0% accuracy, suggesting a good ability to acknowledge uncertainty. It also performs well in "Email Classification (Baseline)" at 99.0% accuracy, indicating strong contextual understanding for categorization tasks. Its "Instruction Following (Baseline)" is respectable at 70.0% accuracy. However, a significant weakness is observed in core cognitive areas such as "Coding," "General Knowledge," "Reasoning," "Ethics," and "Mathematics," where it scored 0.0% accuracy across all these benchmarks. This suggests that while the model excels in speed, cost-efficiency, and certain classification/hallucination avoidance tasks, its capabilities in complex problem-solving, factual recall, and ethical reasoning are currently undeveloped or not adequately captured by these specific tests.