diff --git a/codex-rs/core/src/client.rs b/codex-rs/core/src/client.rs
index 68bd30f4e..f4d830900 100644
--- a/codex-rs/core/src/client.rs
+++ b/codex-rs/core/src/client.rs
@@ -273,6 +273,7 @@ impl ModelClient {
             include,
             prompt_cache_key: Some(self.conversation_id.to_string()),
             text,
+            max_output_tokens: self.config.model_max_output_tokens,
         };
 
         let mut payload_json = serde_json::to_value(&payload)?;
diff --git a/codex-rs/core/src/client_common.rs b/codex-rs/core/src/client_common.rs
index 21f6fc657..e16db3d61 100644
--- a/codex-rs/core/src/client_common.rs
+++ b/codex-rs/core/src/client_common.rs
@@ -280,6 +280,8 @@ pub(crate) struct ResponsesApiRequest<'a> {
     pub(crate) prompt_cache_key: Option<String>,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub(crate) text: Option<TextControls>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub(crate) max_output_tokens: Option<i64>,
 }
 
 pub(crate) mod tools {
@@ -463,6 +465,7 @@ mod tests {
                 verbosity: Some(OpenAiVerbosity::Low),
                 format: None,
             }),
+            max_output_tokens: Some(10_000),
         };
 
         let v = serde_json::to_value(&req).expect("json");
@@ -501,6 +504,7 @@ mod tests {
             include: vec![],
             prompt_cache_key: None,
             text: Some(text_controls),
+            max_output_tokens: Some(10_000),
         };
 
         let v = serde_json::to_value(&req).expect("json");
@@ -537,6 +541,7 @@ mod tests {
             include: vec![],
             prompt_cache_key: None,
             text: None,
+            max_output_tokens: Some(10_000),
         };
 
         let v = serde_json::to_value(&req).expect("json");
diff --git a/codex-rs/core/tests/suite/compact_resume_fork.rs b/codex-rs/core/tests/suite/compact_resume_fork.rs
index f81294baf..cb8258161 100644
--- a/codex-rs/core/tests/suite/compact_resume_fork.rs
+++ b/codex-rs/core/tests/suite/compact_resume_fork.rs
@@ -282,7 +282,8 @@ async fn compact_resume_and_fork_preserve_model_history_view() {
       "include": [
         "reasoning.encrypted_content"
       ],
-      "prompt_cache_key": prompt_cache_key
+      "prompt_cache_key": prompt_cache_key,
+      "max_output_tokens": 128000,
     });
     let compact_1 = json!(
     {
@@ -351,7 +352,8 @@ async fn compact_resume_and_fork_preserve_model_history_view() {
       "include": [
         "reasoning.encrypted_content"
       ],
-      "prompt_cache_key": prompt_cache_key
+      "prompt_cache_key": prompt_cache_key,
+      "max_output_tokens": 128000,
     });
     let user_turn_2_after_compact = json!(
     {
@@ -411,7 +413,8 @@ async fn compact_resume_and_fork_preserve_model_history_view() {
       "include": [
         "reasoning.encrypted_content"
       ],
-      "prompt_cache_key": prompt_cache_key
+      "prompt_cache_key": prompt_cache_key,
+      "max_output_tokens": 128000,
     });
     let usert_turn_3_after_resume = json!(
     {
@@ -491,7 +494,8 @@ async fn compact_resume_and_fork_preserve_model_history_view() {
       "include": [
         "reasoning.encrypted_content"
       ],
-      "prompt_cache_key": prompt_cache_key
+      "prompt_cache_key": prompt_cache_key,
+      "max_output_tokens": 128000,
     });
     let user_turn_3_after_fork = json!(
     {
@@ -571,7 +575,8 @@ async fn compact_resume_and_fork_preserve_model_history_view() {
       "include": [
         "reasoning.encrypted_content"
       ],
-      "prompt_cache_key": fork_prompt_cache_key
+      "prompt_cache_key": fork_prompt_cache_key,
+      "max_output_tokens": 128000,
     });
     let mut expected = json!([
         user_turn_1,