diff --git a/codex-rs/core/BUILD.bazel b/codex-rs/core/BUILD.bazel
index 37a3173f9a..f1649b36c7 100644
--- a/codex-rs/core/BUILD.bazel
+++ b/codex-rs/core/BUILD.bazel
@@ -18,6 +18,7 @@ codex_rust_crate(
     ),
     integration_compile_data_extra = [
         "//codex-rs/apply-patch:apply_patch_tool_instructions.md",
+        "models.json",
         "prompt.md",
     ],
     test_data_extra = [
diff --git a/codex-rs/core/src/models_manager/manager.rs b/codex-rs/core/src/models_manager/manager.rs
index b7cc66159a..7210e1c62d 100644
--- a/codex-rs/core/src/models_manager/manager.rs
+++ b/codex-rs/core/src/models_manager/manager.rs
@@ -239,7 +239,18 @@ impl ModelsManager {
 
     /// Replace the cached remote models and rebuild the derived presets list.
     async fn apply_remote_models(&self, models: Vec<ModelInfo>) {
-        *self.remote_models.write().await = models;
+        let mut existing_models = Self::load_remote_models_from_file().unwrap_or_default();
+        for model in models {
+            if let Some(existing_index) = existing_models
+                .iter()
+                .position(|existing| existing.slug == model.slug)
+            {
+                existing_models[existing_index] = model;
+            } else {
+                existing_models.push(model);
+            }
+        }
+        *self.remote_models.write().await = existing_models;
     }
 
     fn load_remote_models_from_file() -> Result<Vec<ModelInfo>, std::io::Error> {
@@ -272,16 +283,16 @@ impl ModelsManager {
         let chatgpt_mode = self.auth_manager.get_auth_mode() == Some(AuthMode::ChatGPT);
         merged_presets = ModelPreset::filter_by_auth(merged_presets, chatgpt_mode);
 
-        let has_default = merged_presets.iter().any(|preset| preset.is_default);
-        if !has_default {
-            if let Some(default) = merged_presets
-                .iter_mut()
-                .find(|preset| preset.show_in_picker)
-            {
-                default.is_default = true;
-            } else if let Some(default) = merged_presets.first_mut() {
-                default.is_default = true;
-            }
+        for preset in &mut merged_presets {
+            preset.is_default = false;
+        }
+        if let Some(default) = merged_presets
+            .iter_mut()
+            .find(|preset| preset.show_in_picker)
+        {
+            default.is_default = true;
+        } else if let Some(default) = merged_presets.first_mut() {
+            default.is_default = true;
         }
 
         merged_presets
@@ -396,6 +407,16 @@ mod tests {
         .expect("valid model")
     }
 
+    fn assert_models_contain(actual: &[ModelInfo], expected: &[ModelInfo]) {
+        for model in expected {
+            assert!(
+                actual.iter().any(|candidate| candidate.slug == model.slug),
+                "expected model {} in cached list",
+                model.slug
+            );
+        }
+    }
+
     fn provider_for(base_url: String) -> ModelProviderInfo {
         ModelProviderInfo {
             name: "mock".into(),
@@ -415,7 +436,7 @@ mod tests {
     }
 
     #[tokio::test]
-    async fn refresh_available_models_sorts_and_marks_default() {
+    async fn refresh_available_models_sorts_by_priority() {
         let server = MockServer::start().await;
         let remote_models = vec![
             remote_model("priority-low", "Low", 1),
@@ -447,7 +468,7 @@ mod tests {
             .await
             .expect("refresh succeeds");
         let cached_remote = manager.get_remote_models(&config).await;
-        assert_eq!(cached_remote, remote_models);
+        assert_models_contain(&cached_remote, &remote_models);
 
         let available = manager
             .list_models(&config, RefreshStrategy::OnlineIfUncached)
@@ -464,11 +485,6 @@ mod tests {
             high_idx < low_idx,
             "higher priority should be listed before lower priority"
         );
-        assert!(
-            available[high_idx].is_default,
-            "highest priority should be default"
-        );
-        assert!(!available[low_idx].is_default);
         assert_eq!(
             models_mock.requests().len(),
             1,
@@ -508,22 +524,14 @@ mod tests {
             .refresh_available_models(&config, RefreshStrategy::OnlineIfUncached)
             .await
             .expect("first refresh succeeds");
-        assert_eq!(
-            manager.get_remote_models(&config).await,
-            remote_models,
-            "remote cache should store fetched models"
-        );
+        assert_models_contain(&manager.get_remote_models(&config).await, &remote_models);
 
         // Second call should read from cache and avoid the network.
         manager
             .refresh_available_models(&config, RefreshStrategy::OnlineIfUncached)
             .await
             .expect("cached refresh succeeds");
-        assert_eq!(
-            manager.get_remote_models(&config).await,
-            remote_models,
-            "cache path should not mutate stored models"
-        );
+        assert_models_contain(&manager.get_remote_models(&config).await, &remote_models);
         assert_eq!(
             models_mock.requests().len(),
             1,
@@ -587,11 +595,7 @@ mod tests {
             .refresh_available_models(&config, RefreshStrategy::OnlineIfUncached)
             .await
             .expect("second refresh succeeds");
-        assert_eq!(
-            manager.get_remote_models(&config).await,
-            updated_models,
-            "stale cache should trigger refetch"
-        );
+        assert_models_contain(&manager.get_remote_models(&config).await, &updated_models);
         assert_eq!(
             initial_mock.requests().len(),
             1,
diff --git a/codex-rs/core/tests/suite/remote_models.rs b/codex-rs/core/tests/suite/remote_models.rs
index 8a6beb6e24..3b45989d8a 100644
--- a/codex-rs/core/tests/suite/remote_models.rs
+++ b/codex-rs/core/tests/suite/remote_models.rs
@@ -1,4 +1,5 @@
 #![cfg(not(target_os = "windows"))]
+#![allow(clippy::expect_used)]
 // unified exec is not supported on Windows OS
 use std::sync::Arc;
 
@@ -434,8 +435,18 @@ async fn remote_models_preserve_builtin_presets() -> Result<()> {
         .find(|model| model.model == "remote-alpha")
         .expect("remote model should be listed");
     let mut expected_remote: ModelPreset = remote_model.into();
-    expected_remote.is_default = true;
+    expected_remote.is_default = remote.is_default;
     assert_eq!(*remote, expected_remote);
+    let default_model = available
+        .iter()
+        .find(|model| model.show_in_picker)
+        .expect("default model should be set");
+    assert!(default_model.is_default);
+    assert_eq!(
+        available.iter().filter(|model| model.is_default).count(),
+        1,
+        "expected a single default model"
+    );
     assert!(
         available
             .iter()
@@ -451,6 +462,148 @@ async fn remote_models_preserve_builtin_presets() -> Result<()> {
     Ok(())
 }
 
+#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+async fn remote_models_merge_adds_new_high_priority_first() -> Result<()> {
+    skip_if_no_network!(Ok(()));
+    skip_if_sandbox!(Ok(()));
+
+    let server = MockServer::start().await;
+    let remote_model = test_remote_model("remote-top", ModelVisibility::List, -10_000);
+    let models_mock = mount_models_once(
+        &server,
+        ModelsResponse {
+            models: vec![remote_model],
+        },
+    )
+    .await;
+
+    let codex_home = TempDir::new()?;
+    let mut config = load_default_config_for_test(&codex_home).await;
+    config.features.enable(Feature::RemoteModels);
+
+    let auth = CodexAuth::create_dummy_chatgpt_auth_for_testing();
+    let provider = ModelProviderInfo {
+        base_url: Some(format!("{}/v1", server.uri())),
+        ..built_in_model_providers()["openai"].clone()
+    };
+    let manager = ModelsManager::with_provider(
+        codex_home.path().to_path_buf(),
+        codex_core::auth::AuthManager::from_auth_for_testing(auth),
+        provider,
+    );
+
+    let available = manager
+        .list_models(&config, RefreshStrategy::OnlineIfUncached)
+        .await;
+    assert_eq!(
+        available.first().map(|model| model.model.as_str()),
+        Some("remote-top")
+    );
+    assert_eq!(
+        models_mock.requests().len(),
+        1,
+        "expected a single /models request"
+    );
+
+    Ok(())
+}
+
+#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+async fn remote_models_merge_replaces_overlapping_model() -> Result<()> {
+    skip_if_no_network!(Ok(()));
+    skip_if_sandbox!(Ok(()));
+
+    let server = MockServer::start().await;
+    let slug = bundled_model_slug();
+    let mut remote_model = test_remote_model(&slug, ModelVisibility::List, 0);
+    remote_model.display_name = "Overridden".to_string();
+    remote_model.description = Some("Overridden description".to_string());
+    let models_mock = mount_models_once(
+        &server,
+        ModelsResponse {
+            models: vec![remote_model.clone()],
+        },
+    )
+    .await;
+
+    let codex_home = TempDir::new()?;
+    let mut config = load_default_config_for_test(&codex_home).await;
+    config.features.enable(Feature::RemoteModels);
+
+    let auth = CodexAuth::create_dummy_chatgpt_auth_for_testing();
+    let provider = ModelProviderInfo {
+        base_url: Some(format!("{}/v1", server.uri())),
+        ..built_in_model_providers()["openai"].clone()
+    };
+    let manager = ModelsManager::with_provider(
+        codex_home.path().to_path_buf(),
+        codex_core::auth::AuthManager::from_auth_for_testing(auth),
+        provider,
+    );
+
+    let available = manager
+        .list_models(&config, RefreshStrategy::OnlineIfUncached)
+        .await;
+    let overridden = available
+        .iter()
+        .find(|model| model.model == slug)
+        .expect("overlapping model should be listed");
+    assert_eq!(overridden.display_name, remote_model.display_name);
+    assert_eq!(
+        overridden.description,
+        remote_model
+            .description
+            .expect("remote model should include description")
+    );
+    assert_eq!(
+        models_mock.requests().len(),
+        1,
+        "expected a single /models request"
+    );
+
+    Ok(())
+}
+
+#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+async fn remote_models_merge_preserves_bundled_models_on_empty_response() -> Result<()> {
+    skip_if_no_network!(Ok(()));
+    skip_if_sandbox!(Ok(()));
+
+    let server = MockServer::start().await;
+    let models_mock = mount_models_once(&server, ModelsResponse { models: Vec::new() }).await;
+
+    let codex_home = TempDir::new()?;
+    let mut config = load_default_config_for_test(&codex_home).await;
+    config.features.enable(Feature::RemoteModels);
+
+    let auth = CodexAuth::create_dummy_chatgpt_auth_for_testing();
+    let provider = ModelProviderInfo {
+        base_url: Some(format!("{}/v1", server.uri())),
+        ..built_in_model_providers()["openai"].clone()
+    };
+    let manager = ModelsManager::with_provider(
+        codex_home.path().to_path_buf(),
+        codex_core::auth::AuthManager::from_auth_for_testing(auth),
+        provider,
+    );
+
+    let available = manager
+        .list_models(&config, RefreshStrategy::OnlineIfUncached)
+        .await;
+    let bundled_slug = bundled_model_slug();
+    assert!(
+        available.iter().any(|model| model.model == bundled_slug),
+        "bundled models should remain available after empty remote response"
+    );
+    assert_eq!(
+        models_mock.requests().len(),
+        1,
+        "expected a single /models request"
+    );
+
+    Ok(())
+}
+
 #[tokio::test(flavor = "multi_thread", worker_threads = 2)]
 async fn remote_models_request_times_out_after_5s() -> Result<()> {
     skip_if_no_network!(Ok(()));
@@ -588,6 +741,17 @@ async fn wait_for_model_available(
     }
 }
 
+fn bundled_model_slug() -> String {
+    let response: ModelsResponse = serde_json::from_str(include_str!("../../models.json"))
+        .expect("bundled models.json should deserialize");
+    response
+        .models
+        .first()
+        .expect("bundled models.json should include at least one model")
+        .slug
+        .clone()
+}
+
 fn test_remote_model(slug: &str, visibility: ModelVisibility, priority: i32) -> ModelInfo {
     test_remote_model_with_policy(
         slug,