Merge pull request #509 from touwaeriol/pr/antigravity-full

feat(antigravity): comprehensive enhancements - model mapping, rate limiting, scheduling & ops

backend/internal/service/openai_gateway_service.go

@@ -332,7 +332,7 @@ func (s *OpenAIGatewayService) tryStickySessionHit(ctx context.Context, groupID
 
 	// 检查账号是否需要清理粘性会话
 	// Check if sticky session should be cleared
-	if shouldClearStickySession(account) {
+	if shouldClearStickySession(account, requestedModel) {
 		_ = s.cache.DeleteSessionAccountID(ctx, derefGroupID(groupID), cacheKey)
 		return nil
 	}
@@ -498,7 +498,7 @@ func (s *OpenAIGatewayService) SelectAccountWithLoadAwareness(ctx context.Contex
 		if err == nil && accountID > 0 && !isExcluded(accountID) {
 			account, err := s.getSchedulableAccount(ctx, accountID)
 			if err == nil {
-				clearSticky := shouldClearStickySession(account)
+				clearSticky := shouldClearStickySession(account, requestedModel)
 				if clearSticky {
 					_ = s.cache.DeleteSessionAccountID(ctx, derefGroupID(groupID), "openai:"+sessionHash)
 				}