Updated model capabilities across all providers (#940)

app/MindWork AI Studio/Settings/ProviderExtensions.Alibaba.cs

@@ -47,6 +47,43 @@ public static partial class ProviderExtensions
                     Capability.CHAT_COMPLETION_API,
                 ];
             
+            // Check for the Qwen 3.8 family:
+            if(modelName.StartsWith("qwen3.8"))
+            {
+                // Flash thinks by default, but thinking can be turned off:
+                if(modelName.StartsWith("qwen3.8-flash"))
+                    return
+                    [
+                        Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT, Capability.VIDEO_INPUT,
+                        Capability.TEXT_OUTPUT,
+
+                        Capability.REASONING_BY_DEFAULT, Capability.FUNCTION_CALLING,
+                        Capability.CHAT_COMPLETION_API,
+                    ];
+
+                // Unlike the open-weight checkpoint, the Max model keeps its vision
+                // capabilities when used through Alibaba Cloud:
+                if(modelName.StartsWith("qwen3.8-max"))
+                    return
+                    [
+                        Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT, Capability.VIDEO_INPUT,
+                        Capability.TEXT_OUTPUT,
+
+                        Capability.ALWAYS_REASONING, Capability.FUNCTION_CALLING,
+                        Capability.CHAT_COMPLETION_API,
+                    ];
+
+                // All other 3.8 models, such as the 27B one:
+                return
+                [
+                    Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
+                    Capability.TEXT_OUTPUT,
+
+                    Capability.REASONING_BY_DEFAULT, Capability.FUNCTION_CALLING,
+                    Capability.CHAT_COMPLETION_API,
+                ];
+            }
+
             // Check for the 3.0 VL models:
             if(modelName.IndexOf("-vl-") is not -1)
                 return