🐛 fix(provider): add deepseek-v3.1-terminus to THINKING_MODELS (lobehub#9653)

🐛 fix(nvidia): add deepseek-v3.1-terminus to THINKING_MODELS

Add missing deepseek-ai/deepseek-v3.1-terminus model to THINKING_MODELS array
to enable Deep-Think functionality for DeepSeek V3.1 Terminus via NVIDIA NIM.

Fixes lobehub#9648

Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com>
Co-authored-by: Shinji-Li <[email protected]>

Author: ONLY-yours

packages/model-runtime/src/providers/nvidia/index.ts

@@ -3,7 +3,7 @@ import { ModelProvider } from 'model-bank';
 import { createOpenAICompatibleRuntime } from '../../core/openaiCompatibleFactory';
 import { processMultiProviderModelList } from '../../utils/modelParse';
 
-const THINKING_MODELS = ['deepseek-ai/deepseek-v3.1'];
+const THINKING_MODELS = ['deepseek-ai/deepseek-v3.1', 'deepseek-ai/deepseek-v3.1-terminus'];
 
 export interface NvidiaModelCard {
   id: string;