Qwen Plus 0728 (thinking)
qwenqwen-plus-2025-07-28:thinking
Qwen Plus 0728, based on the Qwen3 foundation model, is a 1 million context hybrid reasoning model with a balanced performance, speed, and cost combination.
- Modalities
- text → text
- In / out per 1M
- $0.268 / $0.803
- Context
- 1M tokens
- Added
- Sep 8, 2025
- Max output
- 33K tokens
- Tokenizer
- Qwen3
- Knowledge cutoff
- Mar 31, 2025
- Reasoning
- optional
Pricing
Per 1M tokens. The provider price and our flat 3% fee are separate columns — what you pay is their sum.
| Per 1M tokens | Provider | + 3% fee | You pay |
|---|---|---|---|
| Input | $0.260 | $0.0078 | $0.268 |
| Output | $0.780 | $0.023 | $0.803 |
| Cache write | $0.325 | $0.0097 | $0.335 |
Providers
One company serves Qwen Plus 0728 (thinking). Figures are each provider’s own list price per 1M tokens — your OpenKey price is in the card above.
| Provider | Context | Input /M | Output /M |
|---|---|---|---|
| Alibaba | 1M | $0.260 | $0.780 |
Provider list from the OpenRouter catalog, Jul 28, 2026. All providers
Supported parameters
- include_reasoning
- max_tokens
- presence_penalty
- reasoning
- response_format
- seed
- structured_outputs
- temperature
- tool_choice
- tools
- top_p
Call it
OpenAI-compatible: point your SDK at api.openkey.ai/v1 and use
model qwen/qwen-plus-2025-07-28:thinking.
curl https://api.openkey.ai/v1/chat/completions \
-H "Authorization: Bearer ***" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen/qwen-plus-2025-07-28:thinking",
"messages": [{"role": "user", "content": "Hello"}]
}'import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.openkey.ai/v1",
api_key=os.environ["OPENKEY_API_KEY"],
)
completion = client.chat.completions.create(
model="qwen/qwen-plus-2025-07-28:thinking",
messages=[{"role": "user", "content": "Hello"}],
)
print(completion.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.openkey.ai/v1",
apiKey: process.env.OPENKEY_API_KEY,
});
const completion = await client.chat.completions.create({
model: "qwen/qwen-plus-2025-07-28:thinking",
messages: [{ role: "user", content: "Hello" }],
});
console.log(completion.choices[0].message.content);About Qwen Plus 0728 (thinking)
Qwen Plus 0728 (thinking) is Alibaba's hybrid reasoning variant of the Qwen3-based Plus model, with reasoning that can be toggled per request rather than forced on every call. Its defining spec is a 1 million token context window, one of the largest in the catalog, paired with a 32,768 token completion cap. At $0.26/M input and $0.78/M output on the provider side, it's built for teams that need long-document or long-conversation handling without moving to a frontier-priced model. Text-only in and out, with tool calling and structured outputs supported.
This model sits in the cheaper half of the catalog — 38% of paid models have a lower input price, meaning the median catalog input price ($0.42/M) is over 1.6x what this model charges. Its 1M token context window beats 78% of all catalog models on length. Reasoning is optional here, not mandatory, so you pay for extended thinking only when you request it. It's priced identically to the non-thinking Qwen Plus 0728 sibling on both input and output.
Questions
- How much does Qwen Plus 0728 (thinking) cost via API?
- It costs $0.26 per 1M input tokens and $0.78 per 1M output tokens at provider list price. On OpenKey, the flat 3% fee brings that to $0.2678/M input ($0.26 x 1.03) and $0.8034/M output ($0.78 x 1.03). There's also a $0.325/M cache-write price if you use prompt caching.
- What is Qwen Plus 0728 (thinking)'s context window?
- The context window is 1,000,000 tokens, large enough to hold roughly 1,500 pages of plain text in a single request. Max completion length per response is capped at 32,768 tokens. This puts it ahead of 78% of models in the catalog on context length.
- Does Qwen Plus 0728 (thinking) support tool calling and structured outputs?
- Yes, it supports tool calling (with tool_choice control), structured_outputs, and response_format for JSON-mode style requests. It also exposes reasoning and include_reasoning parameters so you can turn on chain-of-thought reasoning per call rather than having it forced. It does not accept image or other non-text input — it's text-to-text only.
- What is Qwen Plus 0728 (thinking)'s knowledge cutoff?
- Its knowledge cutoff is March 31, 2025. The model itself was made available on the catalog on September 8, 2025.
- How does Qwen Plus 0728 (thinking) compare to Qwen-Plus?
- Qwen Plus 0728 (thinking) shares the exact same pricing as the base Qwen-Plus sibling — $0.26/M input and $0.78/M output on both. The difference is the reasoning toggle: this version adds optional hybrid reasoning support (include_reasoning, reasoning parameters) on top of the same 1M context window and Qwen3 foundation.
More from Qwen
All Qwen models →
QwenQwen3.7 Plus
Qwen3.7-Plus is a cost-effective model in Alibaba's Qwen3.7 series. It supports text and image input with text output, building on the series' text capabilities with a comprehensive upgrade to its...
ctx 1M in $0.330 out $1.32
QwenQwen3.7 Max
Qwen3.7-Max is the flagship model in Alibaba's Qwen3.7 series. It supports text input and output and is designed for agent-centric workloads, with particular strengths in coding, office and productivity tasks,...
ctx 1M in $1.29 out $3.86
QwenQwen3.5 Plus 2026-04-20
Qwen3.5 Plus (April 2026) is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window. This...
ctx 1M in $0.309 out $1.85
QwenQwen3.6 27B
Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026. It features hybrid multimodal capabilities — accepting text, image, and video inputs...
ctx 262K in $0.294 out $2.47