Cache Provider-Modelle in models_cache.json statt bei jedem Start von APIs zu laden

- LoadModelsAsync() prüft zuerst data/models_cache.json
- Bei Cache-Hit: kein API-Call, sofortiger Start
- RefreshModelsFromApis() lädt von APIs + speichert Cache
- POST /api/models/refresh Endpoint für manuellen Refresh
- dataDir an OpenAiService-Konstruktor übergeben

Endpoints/ModelsEndpoint.cs

@@ -1,4 +1,5 @@
 using System.Linq;
+using System.Threading.Tasks;
 using Gott.Services;
 using Microsoft.AspNetCore.Builder;
 using Microsoft.AspNetCore.Http;
@@ -13,6 +14,7 @@ public static class ModelsEndpoint
         app.MapGet("/api/models/nvidia",       HandleNvidiaModels);
         app.MapGet("/api/models/pollinations", HandlePollinationsModels);
         app.MapGet("/api/models/gemini",       HandleGeminiModels);
+        app.MapPost("/api/models/refresh",     HandleRefresh);
     }
 
     private static IResult HandleNvidiaModels(OpenAiService openAiService)
@@ -38,4 +40,10 @@ public static class ModelsEndpoint
             .ToList();
         return Results.Json(new { models });
     }
+
+    private static async Task<IResult> HandleRefresh(OpenAiService openAiService)
+    {
+        await openAiService.RefreshModelsFromApis();
+        return Results.Json(new { success = true, message = "Modelle neu geladen und gecacht." });
+    }
 }