Qwen is Alibaba's family of open models. Its latest generation speaks 119 languages and dialects, comes in a wide range of sizes and has been trained to use MCP tools. Installed on your own hardware, it prepares and sends your campaigns with Mailpro.
Qwen (Tongyi Qianwen) has been developed by Alibaba Cloud, in Hangzhou, China, since 2023. The dense models of its latest generation are published under the Apache 2.0 license, which allows commercial use, from the smallest (under one billion parameters) to several tens of billions.
The Qwen team has announced stronger MCP support and provides a framework, Qwen-Agent, that handles tool calls. For a server like Mailpro's, that is what matters: the model knows how to choose the tool and pass it the right parameters.
Its strength for email marketing: 119 languages and dialects. A company writing to customers in French, German, Italian, Portuguese or beyond can do everything with the same model.
| Developer | Alibaba Cloud, Hangzhou (China) |
|---|---|
| License | Apache 2.0 for the dense models of the latest generation |
| Languages | 119 languages and dialects |
| Sizes | From under one billion to several tens of billions of parameters |
| Tools | Stronger MCP support, Qwen-Agent framework |
Qwen writes each version and sends it to the list in the right language.
A small model on a workstation for simple tasks, a large one on a server for analysis.
Qwen is also good at coding: it helps write the integration that feeds your lists.
When self-hosted, neither your drafts nor your questions go to Alibaba.
Run Qwen with Ollama or LM Studio, in a size that suits your machine.
Chat with it in Open WebUI, LibreChat or LM Studio, or build an agent with Qwen-Agent.
Add the Mailpro MCP server using the server address shown in your Mailpro account, on the AI Assistants (MCP) page, with OAuth sign-in (details: self-hosted AI).
In Mailpro, on the AI Assistants page, set how it sends: with approval or direct, with a daily cap.
When self-hosted, no data goes to Alibaba. If you use Qwen through Alibaba's cloud, apply your usual rules for a provider outside Europe to that service.
The smallest models handle simple requests; to chain several tools (read, write, send), choose a medium or large size.
If an interface handles Qwen's tool calls poorly, Qwen-Agent provides the expected template.
On the Mailpro side, the rules are the same for every assistant: by default, each send waits for your approval, or goes out directly within a daily cap; a log keeps track of everything, and one click suspends all sends. The details.
Yes for the models under Apache 2.0, which is the case for the dense models of the latest generation; check the license of the exact model you choose.
Not if you install Qwen on your own hardware: the model runs on your machines, and the calls only go to your Mailpro account in Switzerland.
A medium size is a good start for chaining reading, writing and sending; the smallest ones are enough for simple tasks.
All questions: AI assistants FAQ.
Create a free Mailpro account, connect your assistant from the AI Assistants page of your account, and stay in control of every send.
