docs: document dynamic model fetching feature

[nil957]

Summary

This PR adds documentation for the dynamic model fetching capability introduced in PR #12046.

Changes

Added a new "Dynamic Model Discovery" section to docs/customize/models.mdx that covers:

1. Supported Providers Table - Shows which providers support auto-loading vs manual refresh:

   • Ollama and OpenRouter auto-load models on startup
   • Anthropic, Gemini, OpenAI, and others support refresh with API key

2. How It Works - Explains the three-step process:

   • Automatic loading for Ollama/OpenRouter
   • Manual refresh using the API key
   • Automatic capability detection

3. Using the Refresh Button - Step-by-step guide for adding models from providers that require API keys

4. Tip - Notes that capabilities (contextLength, maxTokens, tools) are automatically detected and saved

Related Issues

Fixes #12097

Related PR

• feat: fetch provider models dynamically and fix isOpenSource pollution #12046 (feat: fetch provider models dynamically and fix isOpenSource pollution)

---

Summary by cubic

Document dynamic model discovery in docs/customize/models.mdx to explain how models auto-load or refresh across providers. Covers providers that auto-load (Ollama, OpenRouter), those requiring an API key and refresh (Anthropic, Gemini, OpenAI), how to use the refresh button, and that capabilities are auto-detected and saved.

^{Written for commit af574a6. Summary will update on new commits.}

[github-actions]

Thank you for your submission, we really appreciate it. Like many open-source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution. You can sign the CLA by just posting a Pull Request Comment same as the below format.

---

I have read the CLA Document and I hereby sign the CLA

---

Javis seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}

[cubic-dev-ai]

1 issue found across 1 file

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="docs/customize/models.mdx">

<violation number="1" location="docs/customize/models.mdx:52">
P2: Docs describe an internal field (`supportsTools`) as if it were persisted in `config.yaml`, which mismatches the config-facing capability key and can mislead users.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review, or fix all with cubic.}

[cubic-dev-ai]

P2: Docs describe an internal field (supportsTools) as if it were persisted in config.yaml, which mismatches the config-facing capability key and can mislead users.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At docs/customize/models.mdx, line 52:

<comment>Docs describe an internal field (`supportsTools`) as if it were persisted in `config.yaml`, which mismatches the config-facing capability key and can mislead users.</comment>

<file context>
@@ -28,6 +28,44 @@ If no model is specified, Continue automatically uses the default agent with our
+
+1. **Automatic loading**: Ollama and OpenRouter models are automatically loaded when you open Continue
+2. **Manual refresh**: For other providers, enter your API key in the "Add Model" UI and click the refresh icon to fetch available models
+3. **Capability detection**: Continue automatically detects model capabilities (`contextLength`, `maxTokens`, `supportsTools`) and saves them to your `config.yaml`
+
+### Using the Refresh Button
</file context>

Suggested change

	3. **Capability detection**: Continue automatically detects model capabilities (`contextLength`, `maxTokens`, `supportsTools`) and saves them to your `config.yaml`
	3. **Capability detection**: Continue automatically detects model capabilities (for example `contextLength`, `maxTokens`, and tool support via `capabilities.tools`) and saves them to your `config.yaml`