diff --git a/src/cli.rs b/src/cli.rs
index 1b8cbc7..916c0c2 100644
--- a/src/cli.rs
+++ b/src/cli.rs
@@ -86,21 +86,12 @@ impl ChatCLI {
         self.session.add_user_message(message.to_string());
         self.session.save()?;
 
-        // Clone data needed for the API call before getting mutable client reference
-        let model = self.session.model.clone();
-        let messages = self.session.messages.clone();
-        let enable_web_search = self.session.enable_web_search;
-        let enable_reasoning_summary = self.session.enable_reasoning_summary;
-        let reasoning_effort = self.session.reasoning_effort.clone();
-        let enable_extended_thinking = self.session.enable_extended_thinking;
-        let thinking_budget_tokens = self.session.thinking_budget_tokens;
-        
         // Check if we should use streaming before getting client
         let should_use_streaming = {
             let client = self.get_client()?;
             client.supports_streaming()
         };
-        
+
         if should_use_streaming {
             println!(); // Add padding before AI response
             print!("{}> ", console::style("🤖").magenta());
@@ -117,20 +108,20 @@ impl ChatCLI {
                 }) as StreamCallback
             };
             
-            let client = self.get_client()?;
-            match client
+            let client = self.get_client()?.clone();
+            let response = client
                 .chat_completion_stream(
-                    &model,
-                    &messages,
-                    enable_web_search,
-                    enable_reasoning_summary,
-                    &reasoning_effort,
-                    enable_extended_thinking,
-                    thinking_budget_tokens,
+                    &self.session.model,
+                    &self.session.messages,
+                    self.session.enable_web_search,
+                    self.session.enable_reasoning_summary,
+                    &self.session.reasoning_effort,
+                    self.session.enable_extended_thinking,
+                    self.session.thinking_budget_tokens,
                     stream_callback,
                 )
-                .await
-            {
+                .await;
+            match response {
                 Ok(response) => {
                     println!(); // Add newline after streaming
                     self.session.add_assistant_message(response);
@@ -145,20 +136,19 @@ impl ChatCLI {
         } else {
             // Fallback to non-streaming
             let spinner = self.display.show_spinner("Thinking");
-            let client = self.get_client()?;
-            
-            match client
+            let client = self.get_client()?.clone();
+            let response = client
                 .chat_completion(
-                    &model,
-                    &messages,
-                    enable_web_search,
-                    enable_reasoning_summary,
-                    &reasoning_effort,
-                    enable_extended_thinking,
-                    thinking_budget_tokens,
+                    &self.session.model,
+                    &self.session.messages,
+                    self.session.enable_web_search,
+                    self.session.enable_reasoning_summary,
+                    &self.session.reasoning_effort,
+                    self.session.enable_extended_thinking,
+                    self.session.thinking_budget_tokens,
                 )
-                .await
-            {
+                .await;
+            match response {
                 Ok(response) => {
                     spinner.finish("Done");
                     self.display.print_assistant_response(&response);
diff --git a/src/core/client.rs b/src/core/client.rs
index 7db714d..a87846c 100644
--- a/src/core/client.rs
+++ b/src/core/client.rs
@@ -13,7 +13,7 @@ use super::{provider::Provider, session::Message};
 
 pub type StreamCallback = Box<dyn Fn(&str) -> Pin<Box<dyn Future<Output = ()> + Send>> + Send + Sync>;
 
-#[derive(Debug)]
+#[derive(Debug, Clone)]
 pub enum ChatClient {
     OpenAI(OpenAIClient),
     Anthropic(AnthropicClient),
@@ -80,14 +80,14 @@ impl ChatClient {
     }
 }
 
-#[derive(Debug)]
+#[derive(Debug, Clone)]
 pub struct OpenAIClient {
     client: Client,
     api_key: String,
     base_url: String,
 }
 
-#[derive(Debug)]
+#[derive(Debug, Clone)]
 pub struct AnthropicClient {
     client: Client,
     api_key: String,