How to pay for the Qwen API from Russia: a gateway instead of an Alibaba Cloud card

If you need to pay for the Qwen API from Russia, the question is usually not the model but the payment: the official Qwen API is provided by the Alibaba Cloud platform with its own registration rules and list of accepted payment methods. Another path is to top up a Zukko gateway balance via SBP, cryptocurrency or a MasterCard and call Qwen with one key. Below are the steps, prices and a calculation including the fee. Registration and payment terms on Alibaba Cloud's side change, so check the current rules on the service's own site.

Which Qwen models are available and what they cost

Prices as of October 18, 2026, $ per 1M tokens:

For comparison, official prices from the same list: Qwen 3.8 Max input 2 / output 6, Qwen 3.8 Flash 0.15 / 0.47. The strength index is the Intelligence Index from Artificial Analysis. It helps you orient but does not replace testing on your tasks. Before calculating, take fresh Qwen prices from GET https://api.zukko.pro/v1/models or the "Models" page in the dashboard.

Note the cache: for both models the cache read price in the table equals the input price, so there is no separate discount for a repeated prefix. The calculation is simple: input tokens x input price + output tokens x output price.

How to top up the balance

The balance is kept in dollars, top-up from $1. Three methods are available:

• SBP. Payment in rubles at the market rate at the time of payment. The payment page adds an 8.5% fee on top, and the balance receives the dollar amount you requested. Step by step: paying via SBP↗.
• Cryptocurrency (USDT and others): the payment page shows the details.
• MasterCard. A fee of about 13%, amounts from $2 to $594.

An SBP example: to get $10 on the balance you pay the equivalent of 10 x 1.085 = $10.85 in rubles at the rate at the time of payment. A MasterCard example: $10 with a fee of about 13% costs roughly 10 x 1.13 = $11.30. The payment page will show the exact amount before you confirm.

Key and first request

1. Register at zukko.pro↗ and top up the balance.
2. Create a key in the API keys section↗. The full fetch_... key is shown once at creation; save it.
3. Send a request. Qwen is called in the OpenAI format at https://api.zukko.pro/v1:

from openai import OpenAI

client = OpenAI(
    base_url="https://api.zukko.pro/v1",
    api_key="fetch_your_key",
)

resp = client.chat.completions.create(
    model="qwen3.8-flash",
    messages=[
        {"role": "user", "content": "Write three headlines for an article about paying for an API from Russia."}
    ],
    max_tokens=200,
)
print(resp.choices[0].message.content)
print(resp.usage.prompt_tokens, resp.usage.completion_tokens)

The same with curl:

curl https://api.zukko.pro/v1/chat/completions \
  -H "Authorization: Bearer fetch_your_key" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.8-flash","messages":[{"role":"user","content":"Hello"}],"max_tokens":50}'

For a request with an image take qwen3.8-max: in the table it is marked as accepting images, Flash is not. The full path from registration to a request is in the quick start↗.

What working with Qwen costs: a calculation

Say you process 10,000 requests, each with 1,000 input tokens and 300 output tokens. In total that is 10M input tokens and 3M output tokens.

• Qwen 3.8 Flash: 10 x 0.038 + 3 x 0.075 = 0.38 + 0.225 = $0.605
• Qwen 3.8 Max: 10 x 0.088 + 3 x 0.238 = 0.88 + 0.714 = $1.594

Now add paying via SBP. To top the balance up by exactly $2 for a Flash job you pay the equivalent of 2 x 1.085 = $2.17; $2 is also enough for the Max job with $0.41 to spare. For larger top-ups the page fee scales the same way: 8.5% of the top-up amount.

Provider prices change often, so calculate against the current table, not against this article.

Max or Flash: how to choose

Per the table as of October 18, 2026, the difference is this: qwen3.8-max accepts images and has a strength index of 45, qwen3.8-flash works with text only, index 40, and costs noticeably less: in the calculation above the same job came to $0.605 versus $1.594. A sensible order: start with Flash on text tasks (classification, field extraction, short answers), set aside the examples where the result did not satisfy you, and run exactly those through Max. If Max handles them clearly better, use it only for that type of request and keep the rest on Flash. Both models are called with the same key; only the model string changes.

How not to overspend

Cost grows with volume, so it helps to constrain the key. In the dashboard each key has a daily and monthly limit in dollars and a list of allowed models: you can give a script a key for qwen3.8-flash only, with a $1 daily limit. See setting a key limit↗. For bulk tasks, first run 50 examples on Flash and compare against Max on the same examples. A comparison with other Chinese lines is in the article on DeepSeek, Qwen and GLM↗.

FAQ

Do I need an Alibaba Cloud account to use Qwen through Zukko?

No. You work with a Zukko balance and key, and Alibaba Cloud takes no part in your payment or registration.

Which payment method should I pick?

It depends on convenience. SBP adds an 8.5% fee on the payment page, MasterCard about 13%, and cryptocurrency is a separate method. The minimum top-up is $1; compare the final amount on the payment page.

Can I try Qwen for free?

There is no free access. You can get $0.50 for subscribing to the @zukkopro channel (once per Telegram account), and that amount covers thousands of Flash requests if they are short. Check with the formula above.

Does Qwen work only in the OpenAI format?

Any model can be called in both the OpenAI and the Anthropic format. For Qwen an OpenAI-compatible client with the address https://api.zukko.pro/v1 is usually more convenient.