Description
The docker agent setup flow is confusing and the wording appears misleading.
When the setup prompt asks:
How do you want to run models?
1. Cloud provider (needs an API key)
2. Local model via Docker Model Runner (no API key)
3. OpenAI-compatible provider (custom endpoint, e.g. vLLM, LiteLLM)
4. Claude Code harness (Claude subscription via the official `claude` CLI)
it is not clear enough which option should be used for providers like Groq or Hugging Face, and the wording can easily make users think only option 1 is related to API keys while option 3 may also sound like it should be used for other non OpenAI providers.
In practice, this creates confusion about:
- which menu option maps to which provider type
- which key is expected for each provider
- when a provider should be selected under cloud provider vs OpenAI-compatible provider
- whether providers like Groq and Hugging Face should be treated as standard cloud providers or something else
Expected Behavior
The setup flow and documentation should clearly explain:
- option 1 is for built in cloud providers such as OpenAI, Anthropic, Google, Groq, Hugging Face, and others
- option 2 is for local models via Docker Model Runner
- option 3 is only for custom OpenAI-compatible endpoints such as vLLM, LiteLLM, or similar custom base URLs
- option 4 is for Claude Code harness
The wording should make it obvious which option a user should choose before they enter provider selection.
Actual Behavior
The current wording is confusing enough that a user can reasonably misunderstand which option should be selected and what type of API key is expected.
The setup screen shows multiple built in providers after choosing option 1, but the top level prompt and surrounding docs do not make the distinction clear enough.
The first option is unclear because it says "needs an API key" but the flow does not clearly explain how ChatGPT sign-in, Groq, and Hugging Face authentication fit into the setup.
Steps to Reproduce
- Run:
- Read the top level model setup prompt.
- Try to determine which option should be used for providers like Groq or Hugging Face.
- Notice that the wording is not clear enough and can lead to confusion about whether these should be treated as cloud providers or OpenAI-compatible providers.
Docker Agent version
v1.115.0
OS & terminal
Windows 11 - PowerShell
Model used
No response
Error output
This is not a runtime crash. It is a setup flow and documentation clarity issue.
Screenshots
Additional context
This appears to be a UX and documentation issue rather than a provider integration failure.
A clearer setup prompt would reduce confusion for new users and make it easier to choose the correct provider path.
The first option is unclear because it says "needs an API key" but the flow does not clearly explain how ChatGPT sign-in, Groq, and Hugging Face authentication fit into the setup.
Description
The
docker agent setupflow is confusing and the wording appears misleading.When the setup prompt asks:
it is not clear enough which option should be used for providers like Groq or Hugging Face, and the wording can easily make users think only option 1 is related to API keys while option 3 may also sound like it should be used for other non OpenAI providers.
In practice, this creates confusion about:
Expected Behavior
The setup flow and documentation should clearly explain:
The wording should make it obvious which option a user should choose before they enter provider selection.
Actual Behavior
The current wording is confusing enough that a user can reasonably misunderstand which option should be selected and what type of API key is expected.
The setup screen shows multiple built in providers after choosing option 1, but the top level prompt and surrounding docs do not make the distinction clear enough.
The first option is unclear because it says "needs an API key" but the flow does not clearly explain how ChatGPT sign-in, Groq, and Hugging Face authentication fit into the setup.
Steps to Reproduce
Docker Agent version
v1.115.0
OS & terminal
Windows 11 - PowerShell
Model used
No response
Error output
Screenshots
Additional context
This appears to be a UX and documentation issue rather than a provider integration failure.
A clearer setup prompt would reduce confusion for new users and make it easier to choose the correct provider path.
The first option is unclear because it says "needs an API key" but the flow does not clearly explain how ChatGPT sign-in, Groq, and Hugging Face authentication fit into the setup.