diff --git a/packages/app/e2e/helpers/agent-bottom-anchor.ts b/packages/app/e2e/helpers/agent-bottom-anchor.ts
index 4bd8a30235..b4c2cc8293 100644
--- a/packages/app/e2e/helpers/agent-bottom-anchor.ts
+++ b/packages/app/e2e/helpers/agent-bottom-anchor.ts
@@ -131,7 +131,7 @@ export async function seedBottomAnchorAgent(input: {
   const lineCount = Math.max(14, input.lineCount ?? 14);
   const created = await input.client.createAgent({
     provider: "codex",
-    model: "gpt-5.1-codex-mini",
+    model: "gpt-5.4-mini",
     thinkingOptionId: "low",
     modeId: "full-access",
     cwd: input.cwd,
diff --git a/packages/app/e2e/helpers/archive-tab.ts b/packages/app/e2e/helpers/archive-tab.ts
index 5c40d8ee16..5bfe2b9011 100644
--- a/packages/app/e2e/helpers/archive-tab.ts
+++ b/packages/app/e2e/helpers/archive-tab.ts
@@ -109,7 +109,7 @@ export async function createIdleAgent(
 ): Promise<ArchiveTabAgent> {
   const created = await client.createAgent({
     provider: "codex",
-    model: "gpt-5.1-codex-mini",
+    model: "gpt-5.4-mini",
     thinkingOptionId: "low",
     modeId: "full-access",
     cwd: input.cwd,
diff --git a/packages/app/e2e/helpers/daemon-registry.ts b/packages/app/e2e/helpers/daemon-registry.ts
index dd0146890e..b2e7e170d1 100644
--- a/packages/app/e2e/helpers/daemon-registry.ts
+++ b/packages/app/e2e/helpers/daemon-registry.ts
@@ -2,7 +2,7 @@ export const TEST_HOST_LABEL = "localhost";
 
 export const TEST_PROVIDER_PREFERENCES = {
   claude: { model: "haiku" },
-  codex: { model: "gpt-5.1-codex-mini", thinkingOptionId: "low" },
+  codex: { model: "gpt-5.4-mini", thinkingOptionId: "low" },
 } as const;
 
 export function buildDirectTcpConnection(endpoint: string) {
diff --git a/packages/server/src/server/agent/agent-metadata-generator.test.ts b/packages/server/src/server/agent/agent-metadata-generator.test.ts
index f2937de701..ade9543ce9 100644
--- a/packages/server/src/server/agent/agent-metadata-generator.test.ts
+++ b/packages/server/src/server/agent/agent-metadata-generator.test.ts
@@ -11,7 +11,7 @@ import { createAllClients, shutdownProviders } from "./provider-registry.js";
 import { generateAndApplyAgentMetadata } from "./agent-metadata-generator.js";
 import { createWorktree, validateBranchSlug } from "../../utils/worktree.js";
 
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 const shouldRun = !process.env.CI && !!process.env.OPENAI_API_KEY;
diff --git a/packages/server/src/server/agent/agent-response-loop.e2e.test.ts b/packages/server/src/server/agent/agent-response-loop.e2e.test.ts
index f5f0df66ae..da63dae7fe 100644
--- a/packages/server/src/server/agent/agent-response-loop.e2e.test.ts
+++ b/packages/server/src/server/agent/agent-response-loop.e2e.test.ts
@@ -15,7 +15,7 @@ import { createAgentMcpServer } from "./mcp-server.js";
 import { createAllClients, shutdownProviders } from "./provider-registry.js";
 import pino from "pino";
 
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 const hasOpenAICredentials = !!process.env.OPENAI_API_KEY;
diff --git a/packages/server/src/server/agent/agent-response-loop.test.ts b/packages/server/src/server/agent/agent-response-loop.test.ts
index 7920ffbf91..be894c7a55 100644
--- a/packages/server/src/server/agent/agent-response-loop.test.ts
+++ b/packages/server/src/server/agent/agent-response-loop.test.ts
@@ -156,7 +156,7 @@ describe("generateStructuredAgentResponseWithFallback", () => {
       schema,
       providers: [
         { provider: "claude", model: "haiku" },
-        { provider: "codex", model: "gpt-5.1-codex-mini" },
+        { provider: "codex", model: "gpt-5.4-mini" },
       ],
       runner: async (options) => {
         calls.push({
@@ -186,7 +186,7 @@ describe("generateStructuredAgentResponseWithFallback", () => {
       schema,
       providers: [
         { provider: "claude", model: "haiku" },
-        { provider: "codex", model: "gpt-5.1-codex-mini" },
+        { provider: "codex", model: "gpt-5.4-mini" },
       ],
       runner: async (options) => {
         calls.push({
@@ -198,7 +198,7 @@ describe("generateStructuredAgentResponseWithFallback", () => {
     });
 
     expect(result).toEqual({ summary: "ok" });
-    expect(calls).toEqual([{ provider: "codex", model: "gpt-5.1-codex-mini" }]);
+    expect(calls).toEqual([{ provider: "codex", model: "gpt-5.4-mini" }]);
   });
 
   it("falls back when an available provider fails", async () => {
@@ -216,7 +216,7 @@ describe("generateStructuredAgentResponseWithFallback", () => {
       schema,
       providers: [
         { provider: "claude", model: "haiku" },
-        { provider: "codex", model: "gpt-5.1-codex-mini" },
+        { provider: "codex", model: "gpt-5.4-mini" },
       ],
       runner: async (options) => {
         calls.push({
@@ -233,7 +233,7 @@ describe("generateStructuredAgentResponseWithFallback", () => {
     expect(result).toEqual({ summary: "ok" });
     expect(calls).toEqual([
       { provider: "claude", model: "haiku" },
-      { provider: "codex", model: "gpt-5.1-codex-mini" },
+      { provider: "codex", model: "gpt-5.4-mini" },
     ]);
   });
 
@@ -252,7 +252,7 @@ describe("generateStructuredAgentResponseWithFallback", () => {
         schema,
         providers: [
           { provider: "claude", model: "haiku" },
-          { provider: "codex", model: "gpt-5.1-codex-mini" },
+          { provider: "codex", model: "gpt-5.4-mini" },
           { provider: "opencode", model: "opencode/gpt-5-nano" },
         ],
         runner: async () => {
diff --git a/packages/server/src/server/agent/agent-response-loop.ts b/packages/server/src/server/agent/agent-response-loop.ts
index 256d4ab117..d99b2c115b 100644
--- a/packages/server/src/server/agent/agent-response-loop.ts
+++ b/packages/server/src/server/agent/agent-response-loop.ts
@@ -93,7 +93,7 @@ export interface StructuredAgentGenerationWithFallbackOptions<T> {
 
 export const DEFAULT_STRUCTURED_GENERATION_PROVIDERS: readonly StructuredGenerationProvider[] = [
   { provider: "claude", model: "haiku" },
-  { provider: "codex", model: "gpt-5.1-codex-mini", thinkingOptionId: "low" },
+  { provider: "codex", model: "gpt-5.4-mini", thinkingOptionId: "low" },
   { provider: "opencode", model: "opencode/gpt-5-nano" },
 ] as const;
 
diff --git a/packages/server/src/server/agent/provider-manifest.ts b/packages/server/src/server/agent/provider-manifest.ts
index 34bdfd86b1..8d27d2daff 100644
--- a/packages/server/src/server/agent/provider-manifest.ts
+++ b/packages/server/src/server/agent/provider-manifest.ts
@@ -138,7 +138,7 @@ export const AGENT_PROVIDER_DEFINITIONS: AgentProviderDefinition[] = [
     voice: {
       enabled: true,
       defaultModeId: "auto",
-      defaultModel: "gpt-5.1-codex-mini",
+      defaultModel: "gpt-5.4-mini",
     },
   },
   {
diff --git a/packages/server/src/server/agent/providers/opencode-agent.test.ts b/packages/server/src/server/agent/providers/opencode-agent.test.ts
index d31648607b..e6e38f5609 100644
--- a/packages/server/src/server/agent/providers/opencode-agent.test.ts
+++ b/packages/server/src/server/agent/providers/opencode-agent.test.ts
@@ -124,7 +124,7 @@ const hasOpenCode = isBinaryInstalled("opencode");
         m.id.includes("gpt-4.1-nano") ||
         m.id.includes("gpt-4.1-mini") ||
         m.id.includes("gpt-5-nano") ||
-        m.id.includes("gpt-5.1-codex-mini") ||
+        m.id.includes("gpt-5.4-mini") ||
         m.id.includes("gpt-4o-mini"),
     );
 
diff --git a/packages/server/src/server/daemon-e2e/agent-basics.e2e.test.ts b/packages/server/src/server/daemon-e2e/agent-basics.e2e.test.ts
index 50bc048283..a5ea9b8f70 100644
--- a/packages/server/src/server/daemon-e2e/agent-basics.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/agent-basics.e2e.test.ts
@@ -19,8 +19,8 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 describe("daemon E2E", () => {
diff --git a/packages/server/src/server/daemon-e2e/agent-configs.ts b/packages/server/src/server/daemon-e2e/agent-configs.ts
index c54035dac9..ca66c17aef 100644
--- a/packages/server/src/server/daemon-e2e/agent-configs.ts
+++ b/packages/server/src/server/daemon-e2e/agent-configs.ts
@@ -34,7 +34,7 @@ export const agentConfigs = {
   },
   codex: {
     provider: "codex",
-    model: "gpt-5.1-codex-mini",
+    model: "gpt-5.4-mini",
     thinkingOptionId: "low",
     modes: {
       full: "full-access",
diff --git a/packages/server/src/server/daemon-e2e/agent-operations.e2e.test.ts b/packages/server/src/server/daemon-e2e/agent-operations.e2e.test.ts
index 442d6ae03a..6422fce7f0 100644
--- a/packages/server/src/server/daemon-e2e/agent-operations.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/agent-operations.e2e.test.ts
@@ -18,8 +18,8 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 describe("daemon E2E", () => {
diff --git a/packages/server/src/server/daemon-e2e/checkout-ship.e2e.test.ts b/packages/server/src/server/daemon-e2e/checkout-ship.e2e.test.ts
index 25f0dd784f..d9950c5a7e 100644
--- a/packages/server/src/server/daemon-e2e/checkout-ship.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/checkout-ship.e2e.test.ts
@@ -7,7 +7,7 @@ import { execSync } from "child_process";
 import { createDaemonTestContext, type DaemonTestContext } from "../test-utils/index.js";
 import { createWorktree } from "../../utils/worktree.js";
 
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 function tmpCwd(prefix: string): string {
diff --git a/packages/server/src/server/daemon-e2e/file-download.e2e.test.ts b/packages/server/src/server/daemon-e2e/file-download.e2e.test.ts
index dbce142ae8..6fa381b849 100644
--- a/packages/server/src/server/daemon-e2e/file-download.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/file-download.e2e.test.ts
@@ -18,8 +18,8 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 describe("daemon E2E", () => {
diff --git a/packages/server/src/server/daemon-e2e/filesystem.e2e.test.ts b/packages/server/src/server/daemon-e2e/filesystem.e2e.test.ts
index 6294c5512a..73c8290679 100644
--- a/packages/server/src/server/daemon-e2e/filesystem.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/filesystem.e2e.test.ts
@@ -18,8 +18,8 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 describe("daemon E2E", () => {
diff --git a/packages/server/src/server/daemon-e2e/git-operations.e2e.test.ts b/packages/server/src/server/daemon-e2e/git-operations.e2e.test.ts
index 2f5381c86e..f63218a358 100644
--- a/packages/server/src/server/daemon-e2e/git-operations.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/git-operations.e2e.test.ts
@@ -174,8 +174,8 @@ function getWorktreeTerminalBootstrapEntries(
   return terminals as WorktreeTerminalBootstrapEntry[];
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 describe("daemon E2E", () => {
diff --git a/packages/server/src/server/daemon-e2e/images.e2e.test.ts b/packages/server/src/server/daemon-e2e/images.e2e.test.ts
index ef7b33ff7a..461c8d4d3b 100644
--- a/packages/server/src/server/daemon-e2e/images.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/images.e2e.test.ts
@@ -19,7 +19,7 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 
 describe("daemon E2E", () => {
   let ctx: DaemonTestContext;
diff --git a/packages/server/src/server/daemon-e2e/models.e2e.test.ts b/packages/server/src/server/daemon-e2e/models.e2e.test.ts
index dc98d1aa04..24a5e65389 100644
--- a/packages/server/src/server/daemon-e2e/models.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/models.e2e.test.ts
@@ -19,7 +19,7 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 
 function isBinaryInstalled(binary: string): boolean {
   try {
diff --git a/packages/server/src/server/daemon-e2e/permissions-codex.e2e.test.ts b/packages/server/src/server/daemon-e2e/permissions-codex.e2e.test.ts
index d1f424ce31..159191cf46 100644
--- a/packages/server/src/server/daemon-e2e/permissions-codex.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/permissions-codex.e2e.test.ts
@@ -18,8 +18,8 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 describe("daemon E2E", () => {
diff --git a/packages/server/src/server/daemon-e2e/tool-calls.e2e.test.ts b/packages/server/src/server/daemon-e2e/tool-calls.e2e.test.ts
index 51dac983b0..1026812e94 100644
--- a/packages/server/src/server/daemon-e2e/tool-calls.e2e.test.ts
+++ b/packages/server/src/server/daemon-e2e/tool-calls.e2e.test.ts
@@ -19,8 +19,8 @@ function tmpCwd(): string {
   return mkdtempSync(path.join(tmpdir(), "daemon-e2e-"));
 }
 
-// Use gpt-5.1-codex-mini with low thinking preset for faster test execution
-const CODEX_TEST_MODEL = "gpt-5.1-codex-mini";
+// Use gpt-5.4-mini with low thinking preset for faster test execution
+const CODEX_TEST_MODEL = "gpt-5.4-mini";
 const CODEX_TEST_THINKING_OPTION_ID = "low";
 
 type ToolCallItem = Extract<AgentTimelineItem, { type: "tool_call" }>;
diff --git a/packages/server/src/server/persistence-hooks.test.ts b/packages/server/src/server/persistence-hooks.test.ts
index bedcb57348..6a65ce3f74 100644
--- a/packages/server/src/server/persistence-hooks.test.ts
+++ b/packages/server/src/server/persistence-hooks.test.ts
@@ -143,7 +143,7 @@ describe("persistence hooks", () => {
       config: {
         title: "Voice agent (created)",
         modeId: "default",
-        model: "gpt-5.1-codex-mini",
+        model: "gpt-5.4-mini",
         thinkingOptionId: "minimal",
         systemPrompt: "Use speak first.",
         mcpServers: {
@@ -159,7 +159,7 @@ describe("persistence hooks", () => {
     expect(buildConfigOverrides(record)).toMatchObject({
       cwd: "/tmp/project",
       modeId: "plan",
-      model: "gpt-5.1-codex-mini",
+      model: "gpt-5.4-mini",
       thinkingOptionId: "minimal",
       title: "Voice agent (created)",
       systemPrompt: "Use speak first.",
@@ -180,7 +180,7 @@ describe("persistence hooks", () => {
       config: {
         title: "Creation title",
         modeId: "default",
-        model: "gpt-5.1-codex-mini",
+        model: "gpt-5.4-mini",
         systemPrompt: "Confirm and speak first.",
         mcpServers: {
           paseo: {
@@ -196,7 +196,7 @@ describe("persistence hooks", () => {
       provider: "codex",
       cwd: "/tmp/project",
       modeId: "plan",
-      model: "gpt-5.1-codex-mini",
+      model: "gpt-5.4-mini",
       title: "Creation title",
       systemPrompt: "Confirm and speak first.",
       mcpServers: {
diff --git a/packages/server/src/server/speech/providers/local/sherpa/speech-download.e2e.test.ts b/packages/server/src/server/speech/providers/local/sherpa/speech-download.e2e.test.ts
index 08d9ad0df3..15bd4e2237 100644
--- a/packages/server/src/server/speech/providers/local/sherpa/speech-download.e2e.test.ts
+++ b/packages/server/src/server/speech/providers/local/sherpa/speech-download.e2e.test.ts
@@ -202,7 +202,7 @@ describe("speech models (download E2E)", () => {
             provider: "codex",
             cwd: voiceCwd,
             modeId: "full-access",
-            model: "gpt-5.1-codex-mini",
+            model: "gpt-5.4-mini",
             thinkingOptionId: "low",
           },
         });
diff --git a/packages/server/src/server/test-utils/fake-agent-client.ts b/packages/server/src/server/test-utils/fake-agent-client.ts
index f37e174be1..9bd307b374 100644
--- a/packages/server/src/server/test-utils/fake-agent-client.ts
+++ b/packages/server/src/server/test-utils/fake-agent-client.ts
@@ -914,8 +914,8 @@ class FakeAgentClient implements AgentClient {
       return [
         {
           provider: this.provider,
-          id: "gpt-5.1-codex-mini",
-          label: "gpt-5.1-codex-mini",
+          id: "gpt-5.4-mini",
+          label: "gpt-5.4-mini",
           isDefault: true,
         },
       ];
diff --git a/packages/server/src/server/voice-local-agent.e2e.test.ts b/packages/server/src/server/voice-local-agent.e2e.test.ts
index bfd09acefa..c776a2580a 100644
--- a/packages/server/src/server/voice-local-agent.e2e.test.ts
+++ b/packages/server/src/server/voice-local-agent.e2e.test.ts
@@ -54,7 +54,7 @@ function waitForSignal<T>(
       },
       voiceLlmProvider: "codex",
       voiceLlmProviderExplicit: true,
-      voiceLlmModel: "gpt-5.1-codex-mini",
+      voiceLlmModel: "gpt-5.4-mini",
     });
   }, 120000);
 
diff --git a/packages/server/src/server/voice-roundtrip.e2e.test.ts b/packages/server/src/server/voice-roundtrip.e2e.test.ts
index f4e56a5fea..c44d1a870d 100644
--- a/packages/server/src/server/voice-roundtrip.e2e.test.ts
+++ b/packages/server/src/server/voice-roundtrip.e2e.test.ts
@@ -32,7 +32,7 @@ function getVoiceRoundtripConfig(provider: VoiceRoundtripProvider): {
     case "codex":
       return {
         provider: "codex",
-        model: "gpt-5.1-codex-mini",
+        model: "gpt-5.4-mini",
         modeId: "full-access",
         thinkingOptionId: "low",
       };
diff --git a/skills/paseo/SKILL.md b/skills/paseo/SKILL.md
index 312e5916de..61716caa29 100644
--- a/skills/paseo/SKILL.md
+++ b/skills/paseo/SKILL.md
@@ -214,7 +214,7 @@ paseo terminal kill "$id"
 
 **Codex:**
 - `--provider codex/gpt-5.4` — Latest frontier agentic coding model (preferred for all engineering tasks)
-- `--provider codex/gpt-5.1-codex-mini` — Cheaper, faster, but less capable
+- `--provider codex/gpt-5.4-mini` — Cheaper, faster, but less capable
 
 ## Permissions