Mid-tier Qwen with thinking + non-thinking modes and 1M context.
Alibaba documents the qwen-plus family on Bailian with 1M context, thinking mode support, function calling, built-in tools, and structured output. The upstream OpenAI-compatible thinking example uses the non-standard enable_thinking flag in extra_body.
TheRouter serves the model as qwen/qwen-plus on the OpenAI-compatible chat route with catalog context 1,048,576, max output 32,768, and supported parameters including tools, tool_choice, response_format, reasoning, and stop. This reference treats Alibaba's 1M and feature table as upstream authority while calling out TheRouter's parameter spelling and catalog limits.
qwen-plus is the workhorse of the Qwen3 commercial line. It offers two operating modes — fast `enable_thinking: false` for everyday chat and a deeper `enable_thinking: true` mode that emits `reasoning_content` for harder problems — at a fraction of the flagship's price.
The 1M context window makes it a strong default for long-document Q&A, multi-file code reading, and long agent conversations. TheRouter routes between bailian-cn and bailian-sg per request based on cost; `enable_thinking` and `reasoning_content` pass through verbatim in both directions.
| Type | Rate |
|---|---|
| Input | $0.432 / 1M tokens |
| Output | $1.30 / 1M tokens |
| Context length | 1,048,576 tokens in TheRouter catalog; Alibaba documents qwen-plus family context as 1Mhelp.aliyun.com ↗ | verified |
| Maximum output | 32,768 tokens in TheRouter catalog | verified |
| Thinking mode | Supported upstream for the qwen-plus family; OpenAI-compatible calls use enable_thinking upstream, while TheRouter exposes a reasoning parameterhelp.aliyun.com ↗ | verified |
| Function calling and tools | Alibaba's text-generation guide lists function calling for all general models and built-in tools for Plus-tier Qwen recommendationshelp.aliyun.com ↗ | verified |
| Structured output | Supported for qwen-plus family routes in Alibaba's capability tables and exposed by TheRouter as response_formathelp.aliyun.com ↗ | verified |
| Modalities | Text input → text outputhelp.aliyun.com ↗ | verified |
| Benchmark | Distribution | Score | Source |
|---|---|---|---|
Public Plus-tier benchmark table Alibaba's text-generation guide recommends Plus-tier Qwen for balanced capability and cost, but the fetched public page does not publish a numeric benchmark table for the `qwen-plus` alias. | — | Not publicly disclosed | — |
Thinking-mode benchmark The deep-thinking documentation explains how to enable thinking and lists qwen-plus-family support, but does not publish qwen-plus reasoning scores. | — | Not publicly disclosed | — |
TheRouter latency / throughput benchmark No public TTFT or throughput measurement is published for TheRouter's `qwen/qwen-plus` route. Measure with thinking both on and off before setting production SLOs. | — | Not publicly disclosed | — |
Use the global api.therouter.ai endpoint shown below for new integrations; the legacy China accelerated endpoint is retired.
curl https://api.therouter.ai/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer $THE_ROUTER_API_KEY" -d '{
"model": "qwen/qwen-plus",
"messages": [
{"role": "user", "content": "Summarize the key points from this input."}
]
}'Use the OpenAI-compatible chat endpoint for the default non-thinking path. Add reasoning only after validating how your SDK serialises vendor-specific options.
curl https://api.therouter.ai/v1/chat/completions \
-H "Authorization: Bearer $THEROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"qwen/qwen-plus","messages":[{"role":"user","content":"Summarize this customer feedback in three bullets."}],"max_tokens":600}'The current Model Studio guidance recommends Plus-tier Qwen models for office work, chatbots, content generation, document processing, and agent or coding workflows that need 1M context, tools, and balanced cost. The unsuffixed qwen-plus route remains listed as a legacy Qwen alias with the same broad capability columns.
It is TheRouter's OpenAI-compatible route for Alibaba Bailian's commercial qwen-plus alias: a text-in/text-out model with 1M-class context, 32K maximum output, function calling, structured output, and optional reasoning support in the catalog.
Yes, Alibaba's deep-thinking documentation lists qwen-plus-family models among hybrid-thinking routes. Upstream examples use enable_thinking; TheRouter exposes reasoning, so test the exact request body your SDK sends before treating reasoning traces as a stable contract.
Choose qwen-plus for normal production chat, agents, structured output, and document tasks that fit within 1M tokens. Choose qwen-long only when context size is the bottleneck and you accept a narrower long-document route.
No. The public route is an unsuffixed commercial alias. Alibaba separately lists dated Plus snapshots such as Qwen3.7 Plus and older qwen-plus snapshots, so version-specific regressions or benchmark claims should be pinned to a snapshot route when available.
| source | URL | retrieved | |
|---|---|---|---|
| Context length | help.aliyun.com ↗ | 2026-08-02 | verified |
| Maximum output | — | — | verified |
| Thinking mode | help.aliyun.com ↗ | 2026-08-02 | verified |
| Function calling and tools | help.aliyun.com ↗ | 2026-08-02 | verified |
| Structured output | help.aliyun.com ↗ | 2026-08-02 | verified |
| Modalities | help.aliyun.com ↗ | 2026-08-02 | verified |
| Public Plus-tier benchmark table | help.aliyun.com ↗ | 2026-08-02 | unknown |
| Thinking-mode benchmark | help.aliyun.com ↗ | 2026-08-02 | unknown |
| TheRouter latency / throughput benchmark | help.aliyun.com ↗ | 2026-08-02 | unknown |
| Alibaba's text-generation guide positions Plus-tier Qwen as the balanced default | help.aliyun.com ↗ | 2026-08-02 | verified |
| What is qwen/qwen-plus in TheRouter? | help.aliyun.com ↗ | 2026-08-02 | to verify |
| Does qwen-plus support thinking mode? | help.aliyun.com ↗ | 2026-08-02 | to verify |
| When should I choose qwen-plus instead of qwen-long? | help.aliyun.com ↗ | 2026-08-02 | to verify |
| Is qwen-plus a fixed Qwen3.7 checkpoint? | help.aliyun.com ↗ | 2026-08-02 | to verify |