From 2f7f75a20ec0ef3537883e847ff8e44cafb8cf84 Mon Sep 17 00:00:00 2001
From: QuintinShaw <github@xyt.email>
Date: Thu, 6 Aug 2026 04:16:20 +0800
Subject: [PATCH] fix(ai): replace qwen3.8-max-preview with qwen3.8-max on
 token plan (#7670)

GA qwen3.8-max is on models.dev for alibaba-token-plan[-cn]. Point the
Qwen3.8 effort map at the GA id and exclude the retired preview so
regenerate does not revive it.
---
 packages/ai/scripts/generate-models.ts        |  5 ++++-
 .../ai/test/qwen-token-plan-models.test.ts    | 20 +++++++++++++------
 2 files changed, 18 insertions(+), 7 deletions(-)

diff --git a/packages/ai/scripts/generate-models.ts b/packages/ai/scripts/generate-models.ts
index 6c5139993..7dc057f53 100644
--- a/packages/ai/scripts/generate-models.ts
+++ b/packages/ai/scripts/generate-models.ts
@@ -293,6 +293,8 @@ const QWEN_TOKEN_PLAN_REASONING_EFFORT_UNSUPPORTED_MODEL_IDS = new Set([
 	"qwen3.7-max",
 	"qwen3.7-plus",
 ]);
+// Retired preview id — models.dev may still list it after GA ships.
+const QWEN_TOKEN_PLAN_EXCLUDED_MODEL_IDS = new Set(["qwen3.8-max-preview"]);
 
 const KIMI_K3_MAX_TOKENS = 131072;
 const KIMI_K3_COST = {
@@ -2132,6 +2134,7 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
 			for (const [modelId, model] of Object.entries(providerModels)) {
 				const m = model as ModelsDevModel;
 				if (m.tool_call !== true) continue;
+				if (QWEN_TOKEN_PLAN_EXCLUDED_MODEL_IDS.has(modelId)) continue;
 				const supportsReasoningEffort = !QWEN_TOKEN_PLAN_REASONING_EFFORT_UNSUPPORTED_MODEL_IDS.has(modelId);
 
 				models.push({
@@ -2146,7 +2149,7 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
 					...(supportsReasoningEffort
 						? {
 								thinkingLevelMap:
-									modelId === "qwen3.8-max-preview"
+									modelId === "qwen3.8-max"
 										? QWEN_TOKEN_PLAN_QWEN38_THINKING_LEVEL_MAP
 										: QWEN_TOKEN_PLAN_HIGH_MAX_THINKING_LEVEL_MAP,
 							}
diff --git a/packages/ai/test/qwen-token-plan-models.test.ts b/packages/ai/test/qwen-token-plan-models.test.ts
index cf3d9ff1a..98c31cc4a 100644
--- a/packages/ai/test/qwen-token-plan-models.test.ts
+++ b/packages/ai/test/qwen-token-plan-models.test.ts
@@ -53,7 +53,7 @@ const TEXT_MODELS = [
 	"qwen3.6-plus",
 	"qwen3.7-max",
 	"qwen3.7-plus",
-	"qwen3.8-max-preview",
+	"qwen3.8-max",
 ];
 
 const IMAGE_MODELS = ["qwen-image-2.0", "qwen-image-2.0-pro", "wan2.7-image", "wan2.7-image-pro"];
@@ -72,7 +72,7 @@ const QWEN_THINKING_MODELS = [
 	"qwen3.6-plus",
 	"qwen3.7-max",
 	"qwen3.7-plus",
-	"qwen3.8-max-preview",
+	"qwen3.8-max",
 ] as const;
 
 const QWEN_THINKING_MODEL_CASES = (["qwen-token-plan", "qwen-token-plan-cn"] as const).flatMap((provider) =>
@@ -155,9 +155,9 @@ describe("Qwen Token Plan models", () => {
 	it.each(["qwen-token-plan", "qwen-token-plan-cn"] as const)(
 		"exposes qwen3.8 reasoning_effort levels on %s",
 		(provider) => {
-			const model = getModels(provider).find((candidate) => candidate.id === "qwen3.8-max-preview");
+			const model = getModels(provider).find((candidate) => candidate.id === "qwen3.8-max");
 			expect(model).toBeDefined();
-			if (!model) throw new Error(`Missing model: ${provider}/qwen3.8-max-preview`);
+			if (!model) throw new Error(`Missing model: ${provider}/qwen3.8-max`);
 
 			expect(model.thinkingLevelMap).toMatchObject({
 				minimal: null,
@@ -170,6 +170,14 @@ describe("Qwen Token Plan models", () => {
 		},
 	);
 
+	it.each(["qwen-token-plan", "qwen-token-plan-cn"] as const)(
+		"omits retired qwen3.8-max-preview on %s",
+		(provider) => {
+			const modelIds = getModels(provider).map((model) => model.id);
+			expect(modelIds).not.toContain("qwen3.8-max-preview");
+		},
+	);
+
 	it.each(QWEN_REASONING_EFFORT_MODEL_CASES)(
 		"sends Qwen reasoning_effort for $provider/$modelId",
 		async ({ provider, modelId }) => {
@@ -205,9 +213,9 @@ describe("Qwen Token Plan models", () => {
 	it.each(["qwen-token-plan", "qwen-token-plan-cn"] as const)(
 		"sends qwen3.8 max reasoning_effort on %s",
 		async (provider) => {
-			const model = getModels(provider).find((candidate) => candidate.id === "qwen3.8-max-preview");
+			const model = getModels(provider).find((candidate) => candidate.id === "qwen3.8-max");
 			expect(model).toBeDefined();
-			if (!model) throw new Error(`Missing model: ${provider}/qwen3.8-max-preview`);
+			if (!model) throw new Error(`Missing model: ${provider}/qwen3.8-max`);
 
 			let payload: unknown;
 			await streamSimple(