Qwen official rates and deployment regions
Same Alibaba Cloud snapshot as the Chinese page: deployment region and model price, kept apart from a chat subscription. A deployment region is not the buyer’s country.
Figures checked 2026-10-30 and carried over on 2026-11-01. Public list prices can change. Tax, availability, and the final bill follow Alibaba Cloud.
Qwen API: official prices by deployment region
One Qwen3-Max version, same context band, so the regions can be compared. USD per 1 million tokens. A Singapore region can serve approved international workloads. It is not “the price in Singapore for a consumer.”
| Model | Alibaba Cloud region | Input range | Input | Output |
|---|---|---|---|---|
| qwen3-max (2026-01-23) | Singapore | 0–32K tokens | $0.359 | $1.434 |
| qwen3-max (2026-01-23) | Hong Kong | 0–32K tokens | $0.359 | $1.434 |
| qwen3-max (2026-01-23) | Frankfurt | 0–32K tokens | $1.20 | $6.00 |
Real Qwen prices can also change with model version, thinking mode, cache, batch, region, and request length. This table is the comparable official sample, not every region or every model.
Buyer country versus deployment region
The Qwen API is priced by model and deployment region. It is not a “cheaper in Turkey, more expensive in the US” subscription ladder. Service region, available models, and who can pay are separate checks. Beijing, Singapore, and Europe unit prices are not app-store monthly prices.
Chat subscriptions by country
We do not have an official price list that shows the same paid Qwen chat plan for sale in each country. No country monthly fee is filled in.
Ofox next to the official API
Lock the same model, then compare the official input, output, and cache rates with the rates in the Ofox catalog. This page does not type in an Ofox price, and it does not mark any call as smoke-tested. An aggregator can make several models easier to manage. It is not a guaranteed discount, and it may not cover every feature on this page.