Qwen-Flash

Qwen3 Flash，结合思考与非思考模式并支持动态切换。

Alibabaqwen-flash2025-08-01报告数据错误

规格

上下文: 1M
最大输出: 32.8K
输入价: $0.05/ 百万 Token
输出价: $0.4/ 百万 Token
发布日期: 2025-08-01

上下文

上下文: 1M
最大输入: 998K
最大输出: 32.8K

价格 · USD / 百万 Token

输入: $0.05
输出: $0.4
缓存读取: $0.005
缓存写入: $0.063

输入输出

输入: 文本
输出: 文本

API

接口类型: chatmessages

思考

支持思考: 支持
思考等级: —

源信息

状态: 可用
发布日期: 2025-08-01
知识截止: —

能力

已确认: structured_outputstreamingcaching

调用方式

alibabamodel = qwen-flashreasoning = enable_thinking

chat

POSThttps://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1/chat/completions

messages

POSThttps://{WorkspaceId}.ap-southeast-1.maas.aliyuncs.com/apps/anthropic/v1/messages

JSON

qwen-flash.json

{
  "id": "qwen-flash",
  "object": "model",
  "created": 1754006400,
  "owned_by": "alibaba",
  "name": "Qwen-Flash",
  "api": {
    "types": [
      "chat",
      "messages"
    ]
  },
  "limits": {
    "context": 1000000,
    "input": 997952,
    "output": 32768
  },
  "modalities": {
    "input": [
      "text"
    ],
    "output": [
      "text"
    ]
  },
  "reasoning": {
    "supported": true,
    "efforts": []
  },
  "pricing": {
    "currency": "USD",
    "unit": "1M_tokens",
    "input": 0.05,
    "output": 0.4,
    "cache_read": 0.005,
    "cache_write": 0.063
  },
  "features": [
    "structured_output",
    "streaming",
    "caching"
  ],
  "info": {
    "status": "active",
    "release_date": "2025-08-01",
    "knowledge_cutoff": null,
    "description": "Qwen3 Flash combining thinking and non-thinking modes with dynamic switching.",
    "docs": "https://www.alibabacloud.com/help/en/model-studio/text-generation-model",
    "verified_at": "2026-10-11"
  }
}

官方来源

Text generation models | Alibaba Cloud Model Studiowww.alibabacloud.com2026-10-10 Model Studio model pricing | Alibaba Cloudwww.alibabacloud.com2026-10-10 Model lifecycle and updates | Alibaba Cloud Model Studiowww.alibabacloud.com2026-10-11 Anthropic-compatible Messages API | Alibaba Cloud Model Studiowww.alibabacloud.com2026-10-11