What this PR does

Before this PR:
Language detection via LLM used getQuickModel() || getDefaultModel() without validating model availability, and did not disable thinking/reasoning mode. This caused issues when most modern models default to thinking mode, making model selection difficult for language detection.

After this PR:

• Only uses the quick model (falls back to error if unavailable) and validates it exists via hasModel()
• Sets reasoning_effort: 'none' to explicitly disable thinking mode for language detection
• Removes unused getDefaultModel import

Fixes #13562

Why we need it and why it was done in this way

The following tradeoffs were made:

• Removed the getDefaultModel() fallback — language detection should only use the explicitly configured quick model, not silently fall back to the default model which may be slow/expensive.

The following alternatives were considered:

• Passing thinking mode parameters at the API level — rejected because reasoning_effort: 'none' in assistant settings is the existing pattern used elsewhere in the codebase.

Breaking changes

None.

Special notes for your reviewer

Small, focused fix. Only modifies src/renderer/src/utils/translate.ts.

Checklist

• PR: The PR description is expressive enough and will help future contributors
• Code: Write code that humans can understand and Keep it simple
• Refactor: You have left the code cleaner than you found it (Boy Scout Rule)
• Upgrade: Impact of this change on upgrade flows was considered and addressed if required
• Documentation: A user-guide update was considered and is present (link) or not required. Check this only when the PR introduces or changes a user-facing feature or behavior.
• Self-review: I have reviewed my own code (e.g., via /gh-pr-review, gh pr diff, or GitHub UI) before requesting review from others

Release note

fix: Disable thinking/reasoning mode for LLM-based language detection to improve compatibility with modern models that default to thinking mode