Agent Mag
SUBSCRIBE
Qwen3.5-27B
All models
AlibabaAlibabaQwenReleased February 25, 2026

Qwen3.5-27B

262K context$0.195/M input$1.56/M output27B

Qwen3.5-27B is a dense vision-language model that incorporates a linear attention mechanism, optimizing inference speed and performance. It supports reasoning-enabled tasks and structured outputs, making it suitable for complex applications. The model's capabilities are comparable to larger models like Qwen3.5-122B-A10B.

What is Qwen3.5-27B?

Qwen3.5-27B is an AI model from Alibaba that Agent Mag tracks for pricing, context window, modalities, benchmarks, and API compatibility. Builders can use this page to compare Qwen3.5-27B against other models for agent workflows and production deployments.

Model ID

Qwen3.5-27B is a dense vision-language model that incorporates a linear attention mechanism, optimizing inference speed and performance. It supports reasoning-enabled tasks and structured outputs, making it suitable for complex applications. The model's capabilities are comparable to larger models like Qwen3.5-122B-A10B.

Architecture & Specifications
Architecture
Dense Transformer
Parameters
27B
Tokenizer
Qwen3
Released
February 25, 2026
Modalities
Input
textimagevideo
Output
text
Supported Parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
Strengths
  • Supports reasoning-enabled tasks
  • Optimized for fast inference
  • Structured output capabilities
  • Comparable to larger models in performance
  • Vision-language integration
Limitations
  • Limited information on training data
  • Hallucination rate of 20.3%
  • Low accuracy in research-level physics reasoning (0.9%)
  • Moderate coding capability (32.6% on Terminal-Bench Hard)
  • Limited omniscience accuracy (21.0%)
Recommended Use Cases
Vision-language tasks
Scientific reasoning
Instruction-following applications
Agentic coding workflows
Long context reasoning

More from Alibaba

Qwen3.6 Plus
1M ctx$0.325/M

Qwen 3.6 Plus builds on a hybrid architecture that combines efficient linear attention with sparse mixture-of-experts routing, enabling strong scalability and high-performance inference. Compared to the 3.5 series, it delivers...

Qwen3.5-9B
262K ctx$0.100/M

Qwen3.5-9B is a multimodal foundation model from the Qwen3.5 family, designed to deliver strong reasoning, coding, and visual understanding in an efficient 9B-parameter architecture. It uses a unified vision-language design...

Qwen3.5-35B-A3B
262K ctx$0.163/M

The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall...

Qwen3.5-122B-A10B
262K ctx$0.260/M

The Qwen3.5 122B-A10B 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. In terms of...

Related content

Browse the AI models directory

Compare pricing, local installs, context windows, and modality filters across the full model catalog.

Explore AI agent tools

Find frameworks, SDKs, and infrastructure tools that pair with this model in production workflows.

Read Agent Mag model and agent articles

See Agent Mag coverage of model benchmarks, agent frameworks, and deployment patterns.

Data enriched Apr 24, 2026. Pricing from OpenRouter API.