Feature request
Please add first-class support for Alibaba Cloud Model Studio Token Plan's Qwen3.8-Max model to the existing alibaba-coding-plan provider.
Why
Qwen3.8-Max is Alibaba Cloud's current flagship model, advertised for coding and long-horizon agent workflows. Model Studio lists a 1M-token context window and 128K maximum output. It is available through the Token Plan, which is intended to expose Qwen models through OpenAI-compatible APIs.
Jcode already exposes an alibaba-coding-plan provider, but the installed client cannot list its available models until an API key is configured, and Qwen3.8-Max is not surfaced as a documented, first-class selectable model. A native catalog entry and provider validation would make the supported path discoverable and reliable rather than requiring a generic OpenAI-compatible profile.
Proposed acceptance criteria
jcode login alibaba-coding-plan documents the Model Studio / Token Plan credential path and validates the credential without leaking it.jcode model list -p alibaba-coding-plan includes the exact provider model identifier for Qwen3.8-Max after authentication, with a safe static fallback if appropriate.jcode -p alibaba-coding-plan -m <qwen3.8-max-id> supports streaming chat and tool calls through the native provider path.- The model picker presents Qwen3.8-Max with its 1M context window and 128K max-output metadata when the upstream API makes those available.
- Provider doctor or integration coverage includes a live, credential-gated verification for model discovery and a basic streaming/tool-call smoke test.
Scope / non-goals
This request is specifically for the Alibaba Cloud Token Plan / Model Studio native integration. It does not require a new generic OpenAI-compatible provider, and it should not require users to commit or expose their API keys.
References
I am happy to test the supported configuration if maintainers can point to the intended account/API setup.
Feature request
Please add first-class support for Alibaba Cloud Model Studio Token Plan's Qwen3.8-Max model to the existing
alibaba-coding-planprovider.Why
Qwen3.8-Max is Alibaba Cloud's current flagship model, advertised for coding and long-horizon agent workflows. Model Studio lists a 1M-token context window and 128K maximum output. It is available through the Token Plan, which is intended to expose Qwen models through OpenAI-compatible APIs.
Jcode already exposes an
alibaba-coding-planprovider, but the installed client cannot list its available models until an API key is configured, and Qwen3.8-Max is not surfaced as a documented, first-class selectable model. A native catalog entry and provider validation would make the supported path discoverable and reliable rather than requiring a generic OpenAI-compatible profile.Proposed acceptance criteria
jcode login alibaba-coding-plandocuments the Model Studio / Token Plan credential path and validates the credential without leaking it.jcode model list -p alibaba-coding-planincludes the exact provider model identifier for Qwen3.8-Max after authentication, with a safe static fallback if appropriate.jcode -p alibaba-coding-plan -m <qwen3.8-max-id>supports streaming chat and tool calls through the native provider path.Scope / non-goals
This request is specifically for the Alibaba Cloud Token Plan / Model Studio native integration. It does not require a new generic OpenAI-compatible provider, and it should not require users to commit or expose their API keys.
References
I am happy to test the supported configuration if maintainers can point to the intended account/API setup.