Get started
Text model

Qwen by Alibaba — dozens of models, most of them open

Qwen models by Alibaba in GPTunneL: 1M token context, a dedicated branch for code, open weights. Pay per token, no subscription and no VPN.

What the Qwen family is good at

Traits Alibaba keeps from one generation to the next — you can count on them whichever model you pick.

Many sizes

From compact models at eight billion parameters up to the flagship Max tier — you can balance price against quality for the task at hand.

Open weights

Most of the family is published under a free licence: the models can be downloaded and run on your own server.

A branch for code

The separate Coder models are tuned for development — they hold the structure of a project better and edit files more carefully.

Strong Asian languages

The family has been multilingual from the start and handles Chinese and Japanese noticeably better than its competitors.

Which model to pick

The family has three tiers — for different jobs and budgets.

CompactVolume and speedPlusThe default pickMaxHard tasks
Best forClassification, simple extraction, auto-repliesEveryday writing, analysis, codeHard reasoning and large projects
Contextup to 256K tokens1M tokens1M tokens
SpeedHighestHighMedium
PriceLowestLowAbove average

Figures come from the platform catalogue. Exact per-model prices are on the Pricing

Every Qwen model

Release dates follow Alibaba's official announcements.

August 2026

Qwen 3.8 Max

Alibaba's largest model yet: 2.4 trillion parameters, of which only about 95 billion are active on any single answer — flagship quality without the flagship price per token. Handles text, images and video, with a million-token window.

May — June 2026

Qwen 3.7 MaxQwen 3.7 Plus

The previous generation: the flagship Max and the more affordable Plus, both with a million-token window.

April 2026

Qwen 3.6 MaxQwen 3.6 PlusQwen 3.6 27B

A generation focused on compact models — near-flagship quality at a noticeably smaller size.

February 2026

Qwen 3.5 PlusQwen 3.5 FlashQwen 3.5 397B

Context grew to a million tokens for the first time and the fast Flash branch appeared.

2025

Qwen 3 MaxQwen 3 CoderQwen 3 235BQwen 3 30BQwen 3 14BQwen 3 8B

The generation that made people take the family seriously: the flagship Max, a dedicated model for code and a whole set of compact versions.

Full release history
2024

Qwen 2.5 Coder

The generation that introduced a specialised model for development.

How billing works

There is no subscription: you top up one balance and spend it on any model on the platform.

Pay per token

You are charged for exactly what the request and the answer used. Not using it costs nothing: no limits, no monthly fee.

One balance for every model

Qwen, ChatGPT, Claude, image and video generation — all out of the same wallet. No separate subscription per service.

Top up the way you prefer

International cards, Apple Pay, Google Pay or crypto. The minimum top-up is $5.

See pricing

How to get started with Qwen

Sign in to GPTunneL
One account for every model on the platform.
Rewrite this SQL query so it stops doing a full table scan
query.sqlQwen 3.8 Max
Query optimisation

The full scan comes from the function in the WHERE clause — it stops the date index from being used.

Move the conversion to the right-hand side of the comparison and the planner will pick up the index.

A composite index on customer plus date will also remove the sort at the end.

Sign in whichever way suits you.

The model switches right inside the input — you can change it mid-conversation.

The answer lands in the chat and the context is kept until the conversation ends.

Try Qwen in GPTunneL

Signing up takes a minute, and $5 on the balance is enough to see whether the model fits your task.

Frequently asked questions

No. Requests go through GPTunneL's infrastructure, so the models open like any ordinary website — no VPN and no proxy.

Qwen is a family of neural networks from the Chinese company Alibaba and one of the largest in the world by sheer number of models: from compact eight-billion parameter versions to the flagship Max tier. The models answer questions, write and analyse text, program and work with images.

What sets the family apart is the combination of openness and scale: most of the models are published under a free licence and can be downloaded and run on your own server, while the larger versions hold up to a million tokens in context. The family includes a separate Coder branch for development, and Qwen has been strongly multilingual from the start.

Qwen models are available in GPTunneL — worldwide, without a VPN, with one balance and billing for the tokens you actually spend.