diff --git a/packages/core/src/ai-model/common.ts b/packages/core/src/ai-model/common.ts
index f0d67a154..4d721f1d0 100644
--- a/packages/core/src/ai-model/common.ts
+++ b/packages/core/src/ai-model/common.ts
@@ -97,7 +97,7 @@ export function adaptQwenBbox(
 }
 
 export function adaptDoubaoBbox(
-  bbox: number[] | string,
+  bbox: string[] | number[] | string,
   width: number,
   height: number,
 ): [number, number, number, number] {
@@ -127,42 +127,65 @@ export function adaptDoubaoBbox(
     bbox = bbox[0];
   }
 
-  if (bbox.length === 4 || bbox.length === 5) {
+  let bboxList: number[] = [];
+  if (Array.isArray(bbox) && typeof bbox[0] === 'string') {
+    bbox.forEach((item) => {
+      if (typeof item === 'string' && item.includes(',')) {
+        const [x, y] = item.split(',');
+        bboxList.push(Number(x.trim()), Number(y.trim()));
+      } else if (typeof item === 'string' && item.includes(' ')) {
+        const [x, y] = item.split(' ');
+        bboxList.push(Number(x.trim()), Number(y.trim()));
+      } else {
+        bboxList.push(Number(item));
+      }
+    });
+  } else {
+    bboxList = bbox as any;
+  }
+
+  if (bboxList.length === 4 || bboxList.length === 5) {
     return [
-      Math.round((bbox[0] * width) / 1000),
-      Math.round((bbox[1] * height) / 1000),
-      Math.round((bbox[2] * width) / 1000),
-      Math.round((bbox[3] * height) / 1000),
+      Math.round((bboxList[0] * width) / 1000),
+      Math.round((bboxList[1] * height) / 1000),
+      Math.round((bboxList[2] * width) / 1000),
+      Math.round((bboxList[3] * height) / 1000),
     ];
   }
 
   // treat the bbox as a center point
   if (
-    bbox.length === 6 ||
-    bbox.length === 2 ||
-    bbox.length === 3 ||
-    bbox.length === 7
+    bboxList.length === 6 ||
+    bboxList.length === 2 ||
+    bboxList.length === 3 ||
+    bboxList.length === 7
   ) {
     return [
-      Math.max(0, Math.round((bbox[0] * width) / 1000) - defaultBboxSize / 2),
-      Math.max(0, Math.round((bbox[1] * height) / 1000) - defaultBboxSize / 2),
+      Math.max(
+        0,
+        Math.round((bboxList[0] * width) / 1000) - defaultBboxSize / 2,
+      ),
+      Math.max(
+        0,
+        Math.round((bboxList[1] * height) / 1000) - defaultBboxSize / 2,
+      ),
       Math.min(
         width,
-        Math.round((bbox[0] * width) / 1000) + defaultBboxSize / 2,
+        Math.round((bboxList[0] * width) / 1000) + defaultBboxSize / 2,
       ),
       Math.min(
         height,
-        Math.round((bbox[1] * height) / 1000) + defaultBboxSize / 2,
+        Math.round((bboxList[1] * height) / 1000) + defaultBboxSize / 2,
       ),
     ];
   }
 
   if (bbox.length === 8) {
     return [
-      Math.round((bbox[0] * width) / 1000),
-      Math.round((bbox[1] * height) / 1000),
-      Math.round((bbox[4] * width) / 1000),
-      Math.round((bbox[5] * height) / 1000),
+      Math.round((bboxList[0] * width) / 1000),
+      Math.round((bboxList[1] * height) / 1000),
+      Math.round((bboxList[4] * width) / 1000),
+      Math.round((bboxList[5] * height) / 1000),
     ];
   }
 
diff --git a/packages/core/src/ai-model/service-caller/index.ts b/packages/core/src/ai-model/service-caller/index.ts
index aa602df78..69984fa34 100644
--- a/packages/core/src/ai-model/service-caller/index.ts
+++ b/packages/core/src/ai-model/service-caller/index.ts
@@ -31,6 +31,7 @@ import {
   getAIConfig,
   getAIConfigInBoolean,
   getAIConfigInJson,
+  uiTarsModelVersion,
   vlLocateMode,
 } from '@midscene/shared/env';
 import { enableDebug, getDebug } from '@midscene/shared/logger';
@@ -285,7 +286,7 @@ export async function call(
     }
 
     debugProfileStats(
-      `model, ${model}, mode, ${vlLocateMode() || 'default'}, prompt-tokens, ${result.usage?.prompt_tokens || ''}, completion-tokens, ${result.usage?.completion_tokens || ''}, total-tokens, ${result.usage?.total_tokens || ''}, cost-ms, ${Date.now() - startTime}, requestId, ${result._request_id || ''}`,
+      `model, ${model}, mode, ${vlLocateMode() || 'default'}, ui-tars-version, ${uiTarsModelVersion()}, prompt-tokens, ${result.usage?.prompt_tokens || ''}, completion-tokens, ${result.usage?.completion_tokens || ''}, total-tokens, ${result.usage?.total_tokens || ''}, cost-ms, ${Date.now() - startTime}, requestId, ${result._request_id || ''}`,
     );
 
     debugProfileDetail(`model usage detail: ${JSON.stringify(result.usage)}`);
diff --git a/packages/core/tests/unit-test/utils.test.ts b/packages/core/tests/unit-test/utils.test.ts
index 8cb31a00a..4232a48c5 100644
--- a/packages/core/tests/unit-test/utils.test.ts
+++ b/packages/core/tests/unit-test/utils.test.ts
@@ -385,6 +385,30 @@ describe('doubao-vision', () => {
       ]
     `);
   });
+
+  it('adaptDoubaoBbox with string bbox', () => {
+    const result = adaptDoubaoBbox(['123 222', '789 100'], 1000, 2000);
+    expect(result).toMatchInlineSnapshot(`
+      [
+        123,
+        444,
+        789,
+        200,
+      ]
+    `);
+  });
+
+  it('adaptDoubaoBbox with string bbox', () => {
+    const result = adaptDoubaoBbox(['123,222', '789, 100'], 1000, 2000);
+    expect(result).toMatchInlineSnapshot(`
+      [
+        123,
+        444,
+        789,
+        200,
+      ]
+    `);
+  });
 });
 
 describe('doubao-vision', () => {
@@ -415,6 +439,32 @@ describe('doubao-vision', () => {
     const input6 = '123 345 11111';
     const result6 = preprocessDoubaoBboxJson(input6);
     expect(result6).toMatchInlineSnapshot(`"123 345 11111"`);
+
+    const input7 = `
+{
+  "bbox": [
+    "550 216",
+    "550 216",
+    "550 216",
+    "550 216"
+  ],
+  "errors": []
+}
+    `;
+    const result7 = preprocessDoubaoBboxJson(input7);
+    expect(result7).toMatchInlineSnapshot(`
+      "
+      {
+        "bbox": [
+          "550,216",
+          "550,216",
+          "550,216",
+          "550,216"
+        ],
+        "errors": []
+      }
+          "
+    `);
   });
 
   it('adaptDoubaoBbox with 2 points', () => {