From 6bd329701b8025f3c41893c0a5f6a6fed0482e48 Mon Sep 17 00:00:00 2001
From: Gemini AI <gemini@google.com>
Date: Sun, 28 Dec 2025 03:40:56 +0400
Subject: [PATCH] fix: correctly track model provider ID for Qwen and all
 providers

Changes:
1. Added cachedProviderId signal in MultiX v2 component
2. Updated syncFromStore to also sync provider ID from task session
3. Immediately update cached model and provider IDs on model change
4. Pass correct providerId to LiteModelSelector component

This fixes the issue where selecting Qwen models caused chat input to stop
responding, because the provider ID was not being tracked correctly.
---
 packages/ui/src/components/chat/multix-v2/index.tsx | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/packages/ui/src/components/chat/multix-v2/index.tsx b/packages/ui/src/components/chat/multix-v2/index.tsx
index 84f5069..e39f680 100644
--- a/packages/ui/src/components/chat/multix-v2/index.tsx
+++ b/packages/ui/src/components/chat/multix-v2/index.tsx
@@ -69,6 +69,7 @@ export default function MultiXV2(props: MultiXV2Props) {
     const [selectedTaskId, setSelectedTaskIdLocal] = createSignal<string | null>(null);
     const [messageIds, setMessageIds] = createSignal<string[]>([]);
     const [cachedModelId, setCachedModelId] = createSignal("unknown");
+    const [cachedProviderId, setCachedProviderId] = createSignal("");
     const [cachedAgent, setCachedAgent] = createSignal("");
     const [cachedTokensUsed, setCachedTokensUsed] = createSignal(0);
     const [cachedCost, setCachedCost] = createSignal(0);
@@ -184,6 +185,9 @@ export default function MultiXV2(props: MultiXV2Props) {
             if (taskSession?.model?.modelId) {
                 setCachedModelId(taskSession.model.modelId);
             }
+            if (taskSession?.model?.providerId) {
+                setCachedProviderId(taskSession.model.providerId);
+            }
             if (taskSession?.agent) {
                 setCachedAgent(taskSession.agent);
             }
@@ -677,11 +681,14 @@ export default function MultiXV2(props: MultiXV2Props) {
                         <LiteModelSelector
                             instanceId={props.instanceId}
                             sessionId={getActiveTaskSessionId()}
-                            currentModel={{ providerId: "", modelId: cachedModelId() }}
+                            currentModel={{ providerId: cachedProviderId(), modelId: cachedModelId() }}
                             onModelChange={(model) => {
                                 // Update the TASK's session, not a global cache
                                 const taskSessionId = getActiveTaskSessionId();
                                 log.info("[MultiX] Changing model for task session", { taskSessionId, model });
+                                // Immediately update cached values for responsive UI
+                                setCachedModelId(model.modelId);
+                                setCachedProviderId(model.providerId);
                                 updateSessionModelForSession(props.instanceId, taskSessionId, model);
                                 // Force immediate sync to reflect the change
                                 setTimeout(() => syncFromStore(), 50);