fix(gemini): deshabilitar thinking y subir maxOutputTokens a 8192/2048

Con thinkingBudget=0 el modelo no consume tokens en razonamiento interno,
evitando que el JSON de respuesta quede truncado (bug: '{"banco":"Nu",...').
maxOutputTokens 2048→8192 en Vision, 1024→2048 en Intent.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

app/Services/GeminiIntentService.php

@@ -27,7 +27,11 @@ class GeminiIntentService
     {
         return [
             'contents'         => $contents,
-            'generationConfig' => array_merge(['temperature' => 0.1, 'maxOutputTokens' => 1024], $genConfig),
+            'generationConfig' => array_merge([
+                'temperature'    => 0.1,
+                'maxOutputTokens' => 2048,
+                'thinkingConfig' => ['thinkingBudget' => 0],
+            ], $genConfig),
         ];
     }
 

app/Services/GeminiVisionService.php

@@ -34,7 +34,11 @@ class GeminiVisionService
 
         $body  = [
             'contents'         => $contents,
-            'generationConfig' => ['temperature' => 0.1, 'maxOutputTokens' => 2048],
+            'generationConfig' => [
+                'temperature'     => 0.1,
+                'maxOutputTokens' => 8192,
+                'thinkingConfig'  => ['thinkingBudget' => 0],
+            ],
         ];
 
         $inicio    = microtime(true);