Qwen 3.5 Flash
Release Date: 2/23/2026
Benchmarked by 
Fast, cost-effective Qwen 3.5 model with thinking enabled by default; ideal for simple tasks. Supports context cache and batch at half price.
Accuracy (Vals Index)
49.57% ± 1.97
Latency (Vals Index)
316.58s
Cost/Test (Vals Index)
$0.08
Context Window
1M
Max Output Tokens
66k
Input Modality
Hyperparameter settings
Default Provider :
Alibaba
Some benchmarks may use different provider and parameters. Please refer to the benchmark page for more information.
Temperature
1
Top P
Default
Top K
Default
Max Output Tokens
65,536
Benchmarks
Accuracy
Rankings
Academic Benchmarks
Proprietary Benchmarks (contact us to get access)