diff --git a/packages/ai/CHANGELOG.md b/packages/ai/CHANGELOG.md
index da2ea33d..ef945ff3 100644
--- a/packages/ai/CHANGELOG.md
+++ b/packages/ai/CHANGELOG.md
@@ -4,6 +4,7 @@
 
 ### Fixed
 
+- Fixed `openai-codex/gpt-5.3-codex-spark` generated metadata to use its 128k context window ([#4969](https://github.com/earendil-works/pi/issues/4969)).
 - Fixed OpenRouter/Poolside context overflow detection for `maximum allowed input length` errors ([#4943](https://github.com/earendil-works/pi/issues/4943)).
 
 ## [0.75.5] - 2026-05-23
diff --git a/packages/ai/scripts/generate-models.ts b/packages/ai/scripts/generate-models.ts
index e2693a3b..a61813c2 100644
--- a/packages/ai/scripts/generate-models.ts
+++ b/packages/ai/scripts/generate-models.ts
@@ -1512,6 +1512,7 @@ async function generateModels() {
 	// Context window is based on observed server limits (400s above ~272k), not marketing numbers.
 	const CODEX_BASE_URL = "https://chatgpt.com/backend-api";
 	const CODEX_CONTEXT = 272000;
+	const CODEX_SPARK_CONTEXT = 128000;
 	const CODEX_MAX_TOKENS = 128000;
 	const codexModels: Model<"openai-codex-responses">[] = [
 		{
@@ -1547,7 +1548,7 @@ async function generateModels() {
 			reasoning: true,
 			input: ["text"],
 			cost: { input: 1.75, output: 14, cacheRead: 0.175, cacheWrite: 0 },
-			contextWindow: CODEX_CONTEXT,
+			contextWindow: CODEX_SPARK_CONTEXT,
 			maxTokens: CODEX_MAX_TOKENS,
 		},
 		{
diff --git a/packages/ai/src/models.generated.ts b/packages/ai/src/models.generated.ts
index 300ed2c1..efd74010 100644
--- a/packages/ai/src/models.generated.ts
+++ b/packages/ai/src/models.generated.ts
@@ -6977,7 +6977,7 @@ export const MODELS = {
 				cacheRead: 0.175,
 				cacheWrite: 0,
 			},
-			contextWindow: 272000,
+			contextWindow: 128000,
 			maxTokens: 128000,
 		} satisfies Model<"openai-codex-responses">,
 		"gpt-5.4": {