Qwen Plus: API, 1M Context, Pricing & Migration

Qwen Plus, identified as qwen-plus in Alibaba Model Studio, is a legacy hosted text-generation model with a 1M context window. It is a commercial API model, not a downloadable checkpoint intended for local deployment.

Last verified: July 21, 2026.

Important distinction: qwen-plus and qwen3.7-plus are separate API model families. They must not be described as aliases or as the same downloadable model.

Qwen Plus key facts

Provider classification	Legacy Qwen model
Main model ID	`qwen-plus`
Context window	1M tokens
Native input	Text
Output	Text
Access	Alibaba Model Studio hosted API
Public self-host weights	Not published for this identifier
Documented parameter count	Not published for this identifier

Legacy Qwen Plus versus Qwen3.7 Plus

Feature	`qwen-plus`	`qwen3.7-plus`
Provider category	Legacy Qwen	Recommended model
Context	1M	1M
Native modality described on this page	Text	See the separate Qwen3.7 Plus specification
API identifier	`qwen-plus`	`qwen3.7-plus`
Interchangeable	No. Test behavior before changing the model ID.

For new integrations, Alibaba’s recommended-model table places qwen3.7-plus in the balanced performance and cost tier. Existing Qwen Plus applications should migrate only after testing prompts, output structures, tool calls, latency, and costs.

See our separate Qwen3.7 Plus guide for the replacement model’s specification.

Is Qwen Plus downloadable?

No official weight download, parameter count, local deployment recipe, RTX requirement, edge target, mobile target, or open-weight license is published for the qwen-plus API identifier.

Claims that Qwen Plus can be downloaded, deployed on an RTX 3090 or RTX 4090, or embedded on a mobile device confuse this commercial API model with separate open-weight Qwen repositories.

Alias and snapshots

When this page was verified, Alibaba’s pricing table mapped the rolling qwen-plus alias to qwen-plus-2025-12-01. Alibaba also documents identifiers such as qwen-plus-latest and several dated snapshots, with availability varying by deployment scope.

A rolling alias is convenient for provider-managed updates. A dated snapshot offers a more explicit model target where that snapshot is available. Neither identifier should be replaced silently without regression testing.

Thinking, Function Calling, and structured output

Alibaba’s text-model matrix marks qwen-plus and its snapshots as supporting thinking mode, Function Calling, built-in tools, and structured output. Thinking is disabled by default in the documented hybrid-thinking behavior.

Tool support does not mean the model independently executes external actions. Your application must define tools, validate arguments, enforce permissions, execute approved calls, and return tool results to the model.

Qwen Plus pricing

The following standard list prices apply to the Singapore International deployment and were verified against the Alibaba Model Studio pricing document dated July 15, 2026. Prices are per one million tokens and exclude promotional discounts.

Input tokens per request	Input price	Non-thinking output	Thinking output
0–256K	$0.40	$1.20	$4.00
More than 256K–1M	$1.20	$3.60	$12.00

The thinking-output price covers reasoning tokens plus the answer. Regional pricing, promotions, caching, batch pricing, and eligible free quotas can change the final cost, so billing estimates should use the selected deployment scope.

Node.js API example

Use an API key and OpenAI-compatible workspace endpoint from the same Model Studio region.

npm install openai

import OpenAI from "openai";

if (!process.env.DASHSCOPE_API_KEY || !process.env.DASHSCOPE_BASE_URL) {
  throw new Error(
    "Set DASHSCOPE_API_KEY and DASHSCOPE_BASE_URL for the same Model Studio region."
  );
}

const client = new OpenAI({
  apiKey: process.env.DASHSCOPE_API_KEY,
  baseURL: process.env.DASHSCOPE_BASE_URL,
});

const response = await client.chat.completions.create({
  model: "qwen-plus",
  messages: [
    {
      role: "system",
      content: "Be concise and distinguish facts from assumptions.",
    },
    {
      role: "user",
      content: "Create a risk register from the following project notes.",
    },
  ],
  extra_body: {
    enable_thinking: false,
  },
});

console.log(response.choices[0].message.content);

Remove extra_body or configure thinking according to the selected Model Studio endpoint when reasoning output is required. Validate support in the chosen deployment scope before enabling additional API features.

Rate limits

Qwen Plus does not have one universal “queries per minute” value. Alibaba publishes different RPM and TPM limits for rolling aliases, snapshots, deployment scopes, and account configurations. A placeholder such as “X queries per minute” should not appear on a published page.

Production systems should read the scoped rate-limit table, handle HTTP 429 responses, apply bounded retries with jitter, and monitor both request and token throughput.

Migration checklist for Qwen3.7 Plus

Create a test suite from real Qwen Plus requests.
Run the same prompts against qwen3.7-plus.
Compare answer quality, tool-call schemas, structured output, latency, and token usage.
Review modality differences instead of assuming identical input support.
Set cost and rate-limit alerts for the selected deployment scope.
Switch the model ID only after application-level acceptance tests pass.

Frequently asked questions

Is Qwen Plus open source or open weight?

Not under the qwen-plus API identifier. Alibaba has not published downloadable weights or an open-weight license for this hosted commercial model.

Can Qwen Plus run on an RTX 3090 or 4090?

There is no official local deployment package or verified consumer-GPU requirement for qwen-plus. It is accessed through Alibaba Model Studio APIs.

Does Qwen Plus have a known parameter count?

Alibaba does not publish an official parameter count for the hosted qwen-plus identifier.

Is Qwen Plus the same as Qwen3.7 Plus?

No. They are separate model IDs. Qwen Plus appears in the legacy section, while Qwen3.7 Plus appears in Alibaba’s recommended-model table.

Official references

This is an independent technical guide and is not affiliated with or endorsed by Alibaba Cloud or the Qwen team.

Qwen Plus: Hosted API, 1M Context, and Migration