AI model providers · Model provider

Qwen and TestingOps: test agents on your Model Studio key

Add an Alibaba Cloud Model Studio key and the agents on your TestingOps projects run Qwen models through the OpenCode adapter, on your own Alibaba Cloud account.

Model provider
Available with limits

Request a demo Model Studio OpenAI-compatible mode

Qwen AnswerPrompt TestingOpsAI test agents

Connection: Model provider
Writes back: Nothing, only model requests
Last checked: 7 Nov 2026

In short

Qwen and TestingOps in short

The Qwen connection runs TestingOps agents on Qwen models with your own Alibaba Cloud Model Studio key, which the customer admin saves under Integrations → AI models with a preset address for Model Studio's OpenAI-compatible mode.

While Qwen is your first provider, your agents switch to the OpenCode adapter and call Qwen with your key, on qwen-plus unless you choose another model.

Qwen shares one slot with the other OpenAI-compatible providers, and TestingOps' own AI features outside agent runs do not use a Qwen key.

What it does

What it does for your testing

1
Agents on Qwen, billed by Alibaba Cloud
Input
Your Model Studio key, first under Your providers
TestingOps
Switches your organisation's agents to the OpenCode adapter with an alibaba/<model> model and attaches the key to all of your projects
Output
Agent runs on your projects that use Qwen with your key
2
A model name the endpoint lists
Input
The model you enter and the preset address
TestingOps
Test connection sends one short request and checks that the model appears in the endpoint's model list
Output
A warning when the name is not offered, before agents fail on an unknown model
3
One active OpenAI-compatible provider
Input
Qwen together with Kimi, DeepSeek, Azure OpenAI, Ollama or a self-hosted endpoint
TestingOps
Stamps only the highest-ranked of them on your projects and marks the others Standby
Output
A clear view of which of these providers is active

Step by step

How it works

Add Qwen with its preset addressThe customer admin picks Qwen (Alibaba) under Add a model provider. Base URL (preset) holds Model Studio's international OpenAI-compatible address; paste the key and enter a model, or leave the default.
Key and model attached to your projectsTestingOps stores the key encrypted and attaches it to every project of your organisation, together with the address and model.
Qwen at the top drives the agentsWhen Qwen (Alibaba) is marked Default · drives agents, your agents move to the OpenCode adapter at their next run and use Qwen through it.
Only one OpenAI-compatible provider is activeQwen, Kimi, DeepSeek, Azure OpenAI, Ollama and self-hosted endpoints share one slot. Only the highest-ranked of them is stamped on your projects; the others show Standby.

Agent runs reach Qwen through OpenCode's own Alibaba provider. Keys for Model Studio's international site fit the preset address; another regional or workspace address is not confirmed for agent runs yet.

Connect

Setup

Create an API key in Alibaba Cloud Model Studio on its international site.
In TestingOps, as customer admin, open Integrations → AI models and pick Qwen (Alibaba) under Add a model provider.
Keep Base URL (preset), paste the key into API key, enter a Model or leave it empty for the default, and select Save.
Open the entry and select Test connection; fix the model name if TestingOps warns that the endpoint does not list it.
Move Qwen (Alibaba) to the top under Your providers and select Save order, so it shows Default · drives agents instead of Standby.

Documentation: Model Studio OpenAI-compatible mode

Where you connect it in TestingOps: Integrations → AI models → Your providers → Qwen

Use cases

Run test agents on an Alibaba Cloud contract

Trigger: Your company's AI budget sits with Alibaba Cloud Model Studio
Uses: Your Model Studio key, first under Your providers
Agent activity: Agents on your projects write and review cases on Qwen through OpenCode
Result: Agent usage billed to your Alibaba Cloud account

Try a larger Qwen model for long cases

Trigger: Long, multi-step cases need a stronger model than the default
Uses: The Model field of the Qwen entry
Agent activity: Your agents move to the model you enter at their next run
Result: The same projects on a different Qwen model
Condition: Test connection warns if the endpoint does not list that model

Switch between Qwen and Kimi

Trigger: You want to compare two OpenAI-compatible providers
Uses: Both keys under Your providers
Agent activity: Whichever is ranked higher drives the agents; the other waits as Standby
Result: A switch by reordering, without entering keys again
Condition: Save order applies from each agent's next run

Capabilities

Capabilities and limits

What is read

Qwen's responses while agents work on your projects
The endpoint's model list, for Test connection and the Model suggestions

What is produced

Agent runs on your projects that use Qwen through OpenCode
A Reachable or Unreachable mark, and a warning if the model is not listed

Where it can write

Nothing in your Alibaba Cloud account; only model requests are sent with your key.

How it connects

Alibaba Cloud Model Studio in OpenAI-compatible mode with your key; agents run through the OpenCode adapter. A linked RunningOps organisation gets a copy of the key.

Account and permissions

A Model Studio API key
Customer admin rights in TestingOps, or a company admin of the TestingOps workspace managing keys for your organisation

Limits

TestingOps' own AI features, such as case chat or test data suggestions, do not use a Qwen key; they use other models when your package allows it
Only one OpenAI-compatible provider (Qwen, Kimi, DeepSeek, Azure OpenAI, Ollama, self-hosted) is active per organisation; move Qwen to the top to make it the one
The preset address is Model Studio's international OpenAI-compatible endpoint; a different address is not confirmed for agent runs
One key per organisation for all of its projects

Example

Example: how a team uses it

Illustrative scenario, not a customer story.

A team runs most of its AI work on Alibaba Cloud. Its customer admin adds the team's Model Studio key under AI models, keeps the preset address, enters a Qwen model and moves Qwen (Alibaba) to the top of Your providers.

The agents now write the team's cases on Qwen through OpenCode. Case chat in TestingOps still answers on other models, which the admin notes in the team's AI usage review.

Frequently asked questions

Why does Qwen show Standby?

Qwen shares one slot with Kimi, DeepSeek, Azure OpenAI, Ollama and self-hosted endpoints, and only the highest-ranked of them is stamped on your projects. Move Qwen above the others and select Save order.

Do TestingOps' own AI features use our Qwen key?

No. A Qwen key is used for agent runs, the connection test and the copy on RunningOps. Features such as case chat or test data suggestions use other models.

Can we use a Model Studio key from another region?

The preset address is the international OpenAI-compatible endpoint, and agent runs reach Qwen through OpenCode's Alibaba provider. A key or address for another region is not confirmed to work for agent runs yet.

Which model runs if we leave Model empty?

qwen-plus. Enter another model name to move your agents to it; Test connection warns if the endpoint does not list the name.

What happens without a Qwen key?

Your agents do not use your Alibaba Cloud account. They follow your other providers or, if your package allows it, the platform's models.

Next step

Request a demo to see agents on your projects run Qwen with your own Model Studio key.

Request a demo More ai models integrations