商品目录

Qwen vision-language model for visual reasoning, documents, and agent tasks

Qwen vision-language model for visual reasoning, documents, and agent tasks

Qwen vision-language model for visual reasoning, documents, and agent tasks

原站按任务参数或阶梯规则计价，请在原站选择具体规格。

原站按任务参数或阶梯规则计价，请在原站选择具体规格。

The fast, low-overhead option in Qwen3.6. It takes text, image and video input, stretches the context window to a million tokens, and fits when latency and cost are tight.

The fast, low-overhead option in Qwen3.6. It takes text, image and video input, stretches the context window to a million tokens, and fits when latency and cost are tight.

原站按任务参数或阶梯规则计价，请在原站选择具体规格。

原站公开的站内额度标价；充值折扣、规格附加费及最终扣费以原站为准。

Qwen3.6 Flash 是通义千问于 2026-04-27 发布的视觉语言模型，支持文本、图像、视频输入并输出文本，具备推理与工具调用能力。该模型适用于视觉推理、文档理解与智能体任务等场景，支持 1000000 上下文长度与最高 65536 输出长度。

Qwen3.6 Flash 是通义千问于 2026-04-27 发布的视觉语言模型，支持文本、图像、视频输入并输出文本，具备推理与工具调用能力。该模型适用于视觉推理、文档理解与智能体任务等场景，支持 1000000 上下文长度与最高 65536 输出长度。

Qwen 轻量模型。支持文本对话、复杂推理、编码、工具调用与高吞吐轻量调用。适用于高频问答、内容处理与编码辅助。 Qwen lightweight model. Supports text chat, complex reasoning, coding, tool use and high-throughput lightweight calls. Suited for high-frequency Q&A, content processing and coding assistance.

Qwen 轻量模型。支持文本对话、复杂推理、编码、工具调用与高吞吐轻量调用。适用于高频问答、内容处理与编码辅助。 Qwen lightweight model. Supports text chat, complex reasoning, coding, tool use and high-throughput lightweight calls. Suited for high-frequency Q&A, content processing and coding assistance.

原站公开的站内额度标价；充值折扣、规格附加费及最终扣费以原站为准。

原站公开的站内额度标价；充值折扣、规格附加费及最终扣费以原站为准。

原站按任务参数或阶梯规则计价，请在原站选择具体规格。

原站按任务参数或阶梯规则计价，请在原站选择具体规格。

Qwen vision-language model for visual reasoning, documents, and agent tasks

发布日期：2026-04-16 知识截止：官方未公布 更新内容：Qwen3.6 系列高速版：重点提升 agentic coding（多项代码智能体基准大幅超越前代）、数学与代码推理；视觉空间智能、物体定位与目标检测显著增强；1M 上下文。 适合场景：高并发低成本调用、轻量编码与多模态理解、实时应用。

The fast, low-overhead option in Qwen3.6. It takes text, image and video input, stretches the context window to a million tokens, and fits when latency and cost are tight.

The fast, low-overhead option in Qwen3.6. It takes text, image and video input, stretches the context window to a million tokens, and fits when latency and cost are tight.

原站公开的站内额度标价；充值折扣、规格附加费及最终扣费以原站为准。

原站公开的站内额度标价；充值折扣、规格附加费及最终扣费以原站为准。

原站公开的站内额度标价；充值折扣、规格附加费及最终扣费以原站为准。