Qwen4 Max
Neural Network
Alibaba's Qwen4 Max is a flagship Qwen4 model with 1M token context for long documents, images, and video.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 000 000 tokens
- Max answer length
- 393 216 tokens
- Reasoning effort
- none, low, high, extra high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
import OpenAI from 'openai';
const openai = new OpenAI({
apiKey: '<your bothub access token>',
baseURL: 'https://openai.bothub.chat/v1'
});
// Sync - Text generation
async function main() {
const chatCompletion = await openai.chat.completions.create({
messages: [{ role: 'user', content: 'Say this is a test' }],
model: 'qwen4-max',
});
}
// Async - Text generation
async function main() {
const stream = await openai.chat.completions.create({
messages: [{ role: 'user', content: 'Say this is a test' }],
model: 'qwen4-max',
stream: true
});
for await (const chunk of stream) {
const part: string | null = chunk.choices[0].delta?.content ?? null;
}
}
main();
How it works Qwen4 Max?
Qwen4 Max is Alibaba's flagship Qwen4 series model that accepts text, images, and video, outputting text. The 1M token context window allows you to upload entire books, annual reports, or long correspondence and work with them without losing details. Reasoning mode is enabled by default, making the model useful for tasks requiring thoughtful analysis rather than quick answers: data analytics, engineering calculations, legal and financial documents, and video/illustration analysis. Via BotHub, the model is accessible from Russia without a VPN or foreign card: top up your balance with a Russian card in rubles, pay per token, and the balance does not expire. Over 250 models are available in one window alongside Qwen4 Max, allowing you to run the same prompt through multiple models and compare results without separate subscriptions. For product development, there is a unified OpenAI-compatible API that allows switching models without rewriting integrations. Data is encrypted via AES-GCM and not stored.Frequently asked questions about Qwen4 Max
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
qwen4-max works with text, accepts documents and images as input, reasons before answering, and calls functions. A context of up to 1,000,000 tokens allows uploading materials in full, without splitting. Suitable for coding, analytics, and analyzing long reports and documents. In BotHub, it is easy to compare with other models in one window.
According to our data, the maximum response size is 393,216 tokens with a 1,000,000 token context. In practice, this means you can get a voluminous document or a detailed analysis of a large file at once, rather than a short summary. The exact length in characters depends on the language and text formatting.
Yes, reasoning mode is supported—this is noted in the model data. It can break down a task step-by-step, check logic, and then formulate a conclusion, which is useful for mathematics, analytics, and debugging code. If you need a short answer immediately, please request it in your prompt.
Yes, function calling and structured JSON output are supported. The model can be connected to external tools, databases, and APIs, with responses received in a machine-readable format. BotHub provides a unified OpenAI-compatible API for multi-model systems: switching models does not require rewriting integrations.
The model accepts images and documents—this is confirmed in our data, so you can send a scan, file, or image and ask questions about the content. We have no notes on audio and video, so we will not promise their processing. The easiest way to check a specific format is with your own file.
There are no separate quality measurements for Russian in our data, so we will not promise record-breaking results. It is more reliable to test it with your own tasks: send a real request in BotHub, compare responses from several models in one window, and choose the suitable one. Payment is per usage, and tokens do not expire.