diff --git a/inventory-backend/app/services/inbound/base_service.py b/inventory-backend/app/services/inbound/base_service.py
index 0c0625d..25bfaa2 100644
--- a/inventory-backend/app/services/inbound/base_service.py
+++ b/inventory-backend/app/services/inbound/base_service.py
@@ -959,8 +959,12 @@ class MaterialBaseService:
     @staticmethod
     def get_latest_specs():
         """
-        获取所有规格型号的最大连号，按智能分组返回
-        返回格式: [{"group": "S", "latest": "S0115/S0115"}, {"group": "Opt4xxx", "latest": "Opt4018/Opt4018"}, ...]
+        获取所有规格型号的最大连号，按连续区间分组返回
+        - 前缀统一大写处理
+        - 只有数字完全连续（N, N+1, N+2...）才认定为同一组
+        - 数字不连续时断开，形成新组
+        - 按每组数量降序排列
+        - 返回每个连续区间的最大值
         """
         import re
 
@@ -970,55 +974,81 @@ class MaterialBaseService:
             MaterialBase.spec_model != ''
         ).all()
 
-        # 2. 数据结构：{分组名: (原始规格, 数字部分)}
-        groups = {}
+        # 2. 解析并收集所有有效的 (prefix, num, original_spec)
+        parsed = []
 
-        def parse_spec(spec_full):
-            """
-            解析规格型号
-            返回: (prefix, num, group_name, original_spec)
-            """
-            # 取斜杠前的部分作为基准
-            base_spec = spec_full.split('/')[0]
-
-            # 使用正则解析：字母前缀 + 数字
-            match = re.match(r'^([A-Za-z]+)(\d+)$', base_spec)
-            if not match:
-                return (base_spec, 0, base_spec, spec_full)
-
-            prefix, num_str = match.groups()
-            num = int(num_str)
-
-            # 智能分组逻辑
-            if prefix == 'Opt':
-                # Opt 按千位段分组
-                thousand = num // 1000
-                group = f"Opt{thousand}xxx"
-            else:
-                # 常规前缀按原值分组
-                group = prefix
-
-            return (prefix, num, group, spec_full)
-
-        # 3. 遍历并分组
         for material in specs:
             spec = material.spec_model
             if not spec:
                 continue
 
-            prefix, num, group, original_spec = parse_spec(spec)
+            base_spec = spec.split('/')[0]
 
-            if group not in groups:
-                groups[group] = (original_spec, num)
+            match = re.match(r'^([A-Za-z]+)(\d+)$', base_spec)
+            if not match:
+                continue
+
+            prefix, num_str = match.groups()
+            prefix = prefix.upper()
+            num = int(num_str)
+
+            parsed.append((prefix, num, spec))
+
+        # 3. 先按 prefix 升序，再按 num 升序排序
+        parsed.sort(key=lambda x: (x[0], x[1]))
+
+        # 4. 遍历切分连续区间
+        # 核心逻辑：当 current_num != prev_num + 1 时，断开形成新组
+        intervals = []
+        current_prefix = None
+        current_start = None
+        current_end = None
+        current_last_spec = None
+
+        for prefix, num, spec in parsed:
+            if current_prefix is None:
+                current_prefix = prefix
+                current_start = num
+                current_end = num
+                current_last_spec = spec
+            elif prefix == current_prefix and num == current_end + 1:
+                current_end = num
+                current_last_spec = spec
             else:
-                _, existing_num = groups[group]
-                if num > existing_num:
-                    groups[group] = (original_spec, num)
+                intervals.append({
+                    'prefix': current_prefix,
+                    'start': current_start,
+                    'end': current_end,
+                    'count': current_end - current_start + 1,
+                    'latest': current_last_spec
+                })
+                current_prefix = prefix
+                current_start = num
+                current_end = num
+                current_last_spec = spec
 
-        # 4. 构建返回结果，按分组名排序
-        result = [
-            {"group": group, "latest": spec}
-            for group, (spec, _) in sorted(groups.items())
-        ]
+        if current_prefix is not None:
+            intervals.append({
+                'prefix': current_prefix,
+                'start': current_start,
+                'end': current_end,
+                'count': current_end - current_start + 1,
+                'latest': current_last_spec
+            })
+
+        # 5. 按每组数量降序排列，再按前缀升序
+        intervals.sort(key=lambda x: (-x['count'], x['prefix']))
+
+        # 6. 构建返回结果
+        result = []
+        for item in intervals:
+            prefix = item['prefix']
+            start = item['start']
+            end = item['end']
+            result.append({
+                "group": f"{prefix}({start}-{end})",
+                "count": item['count'],
+                "latest": item['latest']
+            })
 
         return result
\ No newline at end of file