fix(vlm): omit unset default max_tokens

[qin-ctx]

Description

Remove the remaining hardcoded VLM max_tokens=32768 fallbacks outside the OpenAI backend. When vlm.max_tokens is not configured, these backends now omit the token-limit parameter and let the selected model/provider apply its own default.

Related Issue

Fixes #2751

Follow-up to #2946, which already removed the OpenAI backend fallback.

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Refactoring (no functional changes)
• Performance improvement
• Test update

Changes Made

• Stop sending max_tokens when vlm.max_tokens is unset in LiteLLM and VolcEngine VLM backends.
• Remove the Kimi-specific default max_tokens constant and fallback.
• Remove the stale Kimi unit-test assertion for the old default.

Testing

• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have tested this on the following platforms:
  • Linux
  • macOS
  • Windows

Not run locally; this is a small request-parameter change and the earlier uv run validation was interrupted before completion.

Checklist

• My code follows the project's coding style
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published

Screenshots (if applicable)

N/A

Additional Notes

Leaving the parameter unset avoids hardcoding one completion-token budget for models and providers with different limits.