Replicating video with body language in API call

Hey Team -

While using the UI I am able to create a video with the correct body language by uploading photo, and using custom motion. Attached is a screenshot of the prompt. (Video ID = f864e70c38aa4f6eba9cb855b4321c21)

However, when I try to recreate this using the API however, I don't get the body language, and the video is very artificial. (VideoID = b1e0793de6e34b8c9c4761d74183a015)

What should I change to ensure that I get a similar video in API as well? Have also specified the API call below.

Curl Request:
curl -X POST https://api.heygen.com/v2/video/generate
-H 'X-Api-Key:
-H 'Content-Type: application/json'
-d '{
"video_inputs": [
{
"character": {
"type": "talking_photo",
"talking_photo_id": "0cb3fa868995495da1153dbcc124be3b",
"avatar_style": "normal"
},
"voice": {
"type": "text",
"input_text": "Hey, I’m Jazeem, founder of Qudemo. We really value every customer, so I wanted to personally welcome you here. Feel free to ask me anything about Qudemo - I’ll walk you through how it works",
"voice_id": "466986b2ee27456bbf9757b7ed72c177",
"speed": 1.1
}
}
],
"dimension": {
"width": 1280,
"height": 720
}
}'

Best,
Sarath