docs: add QVAC to model providers - #12579

Description

Adds a provider page for QVAC, an open-source, local-first runtime that exposes an OpenAI-compatible HTTP server via qvac serve openai. Because the server speaks the OpenAI API, Continue connects to it through the built-in openai provider with a local apiBase — no new provider type is needed.

The page mirrors the existing local / OpenAI-compatible provider docs (e.g. OpenVINO Model Server, vLLM) and covers:

• Installing the CLI (npm i -g @qvac/cli) and starting the server (qvac serve openai, default port 11434).
• Defining model aliases in qvac.config.json and matching the model field to a serve alias.
• A config.yaml / config.json example using provider: openai, apiBase: http://127.0.0.1:11434/v1, and a placeholder apiKey.
• Recommendations: set ctx_size explicitly (the default is small), set reasoning_budget: 0 for reasoning-tuned models like Qwen3, and use a capable agent-tuned model such as gpt-oss-20b for tool-calling quality.

The page is registered in the More Providers navigation and added to the Local Model Options table in the providers overview.

Checklist

• I've read the contributing guide
• The relevant docs, if any, have been updated or created
• The relevant tests, if any, have been updated or created

Screen recording or screenshot

Docs-only change.

Tests

Docs-only change. Verified docs/docs.json remains valid JSON after adding the navigation entry.

Summary by cubic

Add QVAC to the model providers docs with a new page on running its OpenAI-compatible server and connecting via the built-in openai provider using a local apiBase. Updates navigation and the providers overview, switches docs to qvac.tether.io, and adds a note for reasoning-tuned models (e.g., Qwen3.5).

• New Features
  • New QVAC page: install @qvac/cli, start with qvac serve openai (default port 11434).
  • Config examples using provider: openai, apiBase: http://127.0.0.1:11434/v1, any non-empty apiKey, and model aliases from qvac.config.json.
  • Recommendations: set ctx_size; set reasoning_budget: 0 for reasoning-tuned models like Qwen3.5; use a capable model like gpt-oss-20b for tools.
  • Added QVAC to the “More Providers” nav and “Local Model Options” table.

^{Written for commit 2adf2ea. Summary will update on new commits.}