feat(ai): improve model fetching and error handling

- Add server-side model fetching endpoint with flexible provider support
- Refactor client-side AI settings component to use new API query
- Implement dynamic header generation for different AI providers
- Enhance error handling and toast notifications
- Remove local model fetching logic in favor of server-side implementation

packages/server/src/utils/ai/select-ai-provider.ts

@@ -74,8 +74,39 @@ export function selectAIProvider(config: { apiUrl: string; apiKey: string }) {
 				headers: {
 					Authorization: `Bearer ${config.apiKey}`,
 				},
-			)};
+			});
 		default:
 			throw new Error(`Unsupported AI provider: ${providerName}`);
 	}
 }
+
+export const getProviderHeaders = (
+	apiUrl: string,
+	apiKey: string,
+): Record<string, string> => {
+	// Anthropic
+	if (apiUrl.includes("anthropic")) {
+		return {
+			"x-api-key": apiKey,
+			"anthropic-version": "2023-06-01",
+		};
+	}
+
+	// Mistral
+	if (apiUrl.includes("mistral")) {
+		return {
+			Authorization: apiKey,
+		};
+	}
+
+	// Default (OpenAI style)
+	return {
+		Authorization: `Bearer ${apiKey}`,
+	};
+};
+export interface Model {
+	id: string;
+	object: string;
+	created: number;
+	owned_by: string;
+}