MODEL PROFILE · 模型档案

Qwen3.5-2B

阿里巴巴 · Qwen · Qwen3.5-2B

> [!Note] > This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. > > These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, KTransformers, etc. > > In light of its parameter scale, the intended use cases are prototyping, task-specific fine-tuning, and other research or development purposes. Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance. Qwen3.5 represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility to empower developers and enterprises with unprecedented capability and efficiency. - **Unified Vision-Language Foundation**: Early fusion training on multimodal tokens achieves cross-generational parity with Qwen3 and outperforms Qwen3-VL models across reasoning, coding, agents, and visual understanding benchmarks.

译文排队中,当前展示来源原文。

文本生成license:apache-2.0
上下文窗口262,144 Token
输入 → 输出文本 → 文本
模型权重开放权重
档案更新时间2026/11/12 08:41:23
OVERVIEW

模型介绍

> [!Note] > This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. > > These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, KTransformers, etc. > > In light of its parameter scale, the intended use cases are prototyping, task-specific fine-tuning, and other research or development purposes. Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance. Qwen3.5 represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility to empower developers and enterprises with unprecedented capability and efficiency. - **Unified Vision-Language Foundation**: Early fusion training on multimodal tokens achieves cross-generational parity with Qwen3 and outperforms Qwen3-VL models across reasoning, coding, agents, and visual understanding benchmarks.

下方摘要由共用资料处理模型依据来源整理,非本站实测;选型结论请结合原始资料和实际业务验证。

主要特点

  • 支持Hugging Face Transformers、vLLM、SGLang、KTransformers等主流推理框架
  • 专为原型开发、任务特定微调及研究开发场景优化的2B参数规模模型
  • 采用早期融合(early fusion)训练的统一视觉-语言基础架构,覆盖文本与视觉模态

局限与注意事项

暂无足够来源支持具体缺点结论。请重点验证实际任务表现、时延、数据政策、许可与版本生命周期。

PRICING & COST

来源报价与费用分析

同一模型在不同渠道、地区和配置下可能采用不同价格。下表独立保留各渠道报价,不将渠道价格视为厂商官方统一价格。

来源 / 渠道输入输出计费单位采集时间
暂未取得可核验报价,请查阅来源渠道。

按月估算基础 Token 费用

—

预估不包含缓存优惠、阶梯定价、多模态附加费用、网络检索、税费及推理额外输出。

INDEPENDENT BENCHMARKS

公开评测与运行配置

暂无已采集的可核验公开评分。模型未上榜不代表能力较低。

查看完整已收录榜单 →

常见问题

Qwen3.5-2B是否支持多模态输入?

是的,基于统一视觉-语言基础架构,支持 multimodal tokens 输入,具备视觉理解能力。

该模型能否直接用于生产环境部署?

官方明确指出其定位为原型开发与研究用途,未声明适用于直接生产部署。

它与Qwen3-VL模型在视觉理解任务上的关系是什么?

在视觉理解、推理、代码及智能体等基准上均优于Qwen3-VL模型。

开放权重是否代表可以免费商用?

不一定。应核对具体版本许可证、商用门槛和使用政策。自部署还会产生算力和维护成本。

这些价格是实时结算价格吗?

不是实时结算价格,而是带采集时间的来源报价快照。请到对应渠道核对当前版本、区域、折扣和计费方式。

资料来源与可追溯记录

许可证:apache-2.0。模型介绍与第三方评测均以原来源的具体版本和方法为准。