The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers state-of-the-art performance comparable to leading-edge models across a wide range of tasks, including language understanding, logical reasoning, code generation, agent-based tasks, image understanding, video understanding, and graphical user interface (GUI) interactions. With its robust code-generation and agent capabilities, the model exhibits strong generalization across diverse agent.
Evaluations
13
across 13 benchmarks
Latency
4m 13s
Context length
262k
tokens
Cost
$0.60 · $3.60
input · output per 1M tokens
Key takeaways
Qwen3.5 397B A17B is a multimodal vision-language model designed for diverse tasks including language, image, and video understanding, logical reasoning, code generation, and GUI interactions, aiming to solve complex multi-domain problems.
The model utilizes a hybrid architecture combining linear attention with a sparse Mixture-of-Experts (MoE) to achieve higher inference efficiency and state-of-the-art performance across various domains.
Its notable capabilities include robust code generation, strong generalization across diverse agent tasks, and exceptional problem-solving in advanced mathematics, as evidenced by 100% accuracy on the AIME 2024 benchmark. It also demonstrates proficiency in addressing logic-based errors in symbolic manipulation.
The model struggles with complex symbolic evaluation, particularly with LaTeX parsing (20% on SWE-bench Lite) due to issues with precedence and grouping, handling symbolic 'i' in complex expressions, nuanced interpretations of operator precedence, and consistently handling parsing in 'evaluate=False' contexts.
Qwen3.5 achieved perfect accuracy (100%) on the AIME 2024 benchmark, demonstrating strong mathematical reasoning and handling diverse advanced mathematical topics, but scored 20% on SWE-bench Lite, highlighting challenges in complex symbolic evaluation and parsing.
