Venice not setting provider transform for Minimax or Kimi, or small options.

[dpuyosa]

Description

Venice has different naming convention for some models so they don't get proper temperature, topP and topK.
Also, when running small model, it is using unnecessary full reasoning.

Plugins

No response

OpenCode version

v1.1.65

Steps to reproduce

With Venice, use Minimax M2.1 or M2.5, Kimi K2.5 -> Transforms are not set.
Use any Venice model as small -> Full default reasoning.

Screenshot and/or share link

No response

Operating System

No response

Terminal

No response

[github-actions]

This issue might be a duplicate of or related to existing issues. Please check:

• Subtle bug: model.variants can be unintentionally overwritten when ProviderTransform.variants() returns {} #12169: Subtle bug with model.variants and ProviderTransform.variants() - directly related to provider transform mechanism
• missing reasoning variants for multiple providers #12155: Missing reasoning variants for multiple providers
• Models from venice provider doesn't work #9108: Models from venice provider doesn't work (though with different error symptoms)

These may help provide additional context or might already have solutions in progress.