RU

Home / APIs & tokens / Qwen / Subscriptions

Qwen subscription

Alibaba Cloud

—

No offers yet

Nobody has listed this product yet. We do not invent prices: a displayed price is a public offer, and it only ever comes from a live seller.

Log in to get the notification: without an account there is nowhere to send it.

Plans

Pay-as-you-go

  • Billed per token for actual usage — input and output counted separately, no monthly quota and no commitment
  • Qwen3.8-Max: context window up to 1M tokens, up to 131,072 output tokens, thinking mode, function calling and structured outputs
  • Text models Qwen3.7-Max, Qwen3.7-Plus, Qwen3.7-Flash and Qwen3.6-Flash, plus the Qwen3.5-Omni-Plus omni model and its realtime version
  • Coding models Qwen3-Coder-Next and Qwen3-Coder-Plus, image models Qwen-Image-2.0-Pro and Wan2.7-Image-Pro, text-embedding-v4 embeddings and qwen3-rerank reranking
  • The same key also unlocks third-party DeepSeek-V4-Pro, DeepSeek-V4-Flash, GLM-5.2, Kimi-K2.7-Code and MiniMax-M2.5
  • Context caching and built-in web search (Beijing and Singapore regions); the API is OpenAI-compatible and a DashScope SDK is available

Coding Plan Pro

  • 6,000 requests per five-hour window, 45,000 per week and 90,000 per month
  • Plan models: qwen3.7-plus, qwen3.6-plus, kimi-k2.5, glm-5 and MiniMax-M2.5, plus qwen3.5-plus, qwen3-max-2026-01-23, qwen3-coder-next, qwen3-coder-plus and glm-4.7
  • Works in Claude Code, Cursor, Cline, Qwen Code, Codex, OpenCode, OpenClaw, Kilo CLI, Lingma, Qoder, Cherry Studio and more — over a dozen tools
  • A dedicated sk-sp-… key and OpenAI- and Anthropic-compatible endpoints; the plan is meant for interactive editor work, not scripts or server-side integrations
  • The monthly quota resets on your renewal date, the weekly one every Monday, and the five-hour one runs as a rolling window
  • A single editor request spends several model calls: a simple task usually 5–10, a complex one 10–30 or more

Plan contents as published by the vendor; seller prices arrive at launch.

What you get

A Qwen Coding Plan subscription is issued on your own Alibaba Cloud account: you get a dedicated sk-sp-… key and paste it into Claude Code, Cursor, Cline, Qwen Code or another supported tool. Check the plan — it determines how many requests you get per five hours, per week and per month, and keep in mind that one editor task spends several model calls at once. Remember the vendor restriction: the plan key is meant for interactive use in development tools, not for scripts or server-side integrations.

About the service

Access to Qwen models through Alibaba Cloud Model Studio: billed per token for the volume you actually use, with an OpenAI-compatible key. The flagship is Qwen3.8-Max, released on 3 August 2026: 2.4 trillion parameters with 95 billion active, a context window of up to 1M tokens, thinking mode, tool calling, structured outputs and understanding of images and video. Alongside it sit Qwen3.7-Max, Qwen3.7-Plus, Qwen3.7-Flash and Qwen3.6-Flash, the realtime omni model Qwen3.5-Omni-Plus, the coding models Qwen3-Coder-Next and Qwen3-Coder-Plus, image generation with Qwen-Image-2.0-Pro and Wan2.7-Image-Pro, plus third-party DeepSeek-V4-Pro, DeepSeek-V4-Flash, GLM-5.2, Kimi-K2.7-Code and MiniMax-M2.5 on the same platform. Separately from tokens the vendor sells the Qwen Coding Plan subscription: a fixed monthly request quota instead of a usage bill, usable straight inside Claude Code, Cursor, Cline, Qwen Code and a dozen more editors and agents. New Model Studio users get a free trial allowance of roughly 1,000,000 tokens per model, valid for 90 days.