Skip to main content
Guild supports Qwen models from Alibaba. This page lists the models Guild recognizes, the inference providers that serve them, and how to call them from an agent.

​
Inference providers

Qwen models are served only through third-party inference providers, not a Qwen API. Write the canonical model name, such as qwen3-next-80b-a3b, everywhere in Guild. Guild maps it to each provider’s own model ID.

​
Models

Guild prices and recognizes these Qwen model IDs. A policy or preference can also name any other model your credential can reach.

​
Set up Qwen

Guild has no managed key for Qwen. To use Qwen models, add a credential for a provider that serves them. Adding your first credential switches the whole account to bring your own key (BYOK).
1

Open Models & providers

Click Access & setup in the left nav, then Models & providers. In an organization, only admins can open it.
2

Add a credential

Click Add API key and choose an inference provider from the table above.
3

Enter the provider's secret

Give the credential a Name and paste your provider’s API key. AWS Bedrock asks for an IAM role and region instead.
4

Choose Alibaba

Under Models, check Alibaba. Keep qwen3-next-80b-a3b as its default model or enter another from the list above. Then click Add Key.
To control which workspaces and agents can use Qwen models, add a model policy for the credential. For example, --models "qwen3-coder-*" limits a policy to matching models:

​
Use it from an agent

Agent code never names a credential. Guild resolves the call against the workspace owner’s settings. To ask for Qwen explicitly, put it in llmPreferences:
The first entry asks for qwen3-coder-480b-a35b. The second falls back to the default model of the matching policy. Preferences are strict: if no credential or policy allows any of them, the call fails rather than using another publisher. See LLM preferences.
The "alibaba" provider value needs @guildai/agents-sdk 0.7.6 or later.
  • Models & providers for credentials, model policies, and the daily token limit
  • LLMs for task.llm and the unified LLM proxy
  • Usage for spend by model