Enable image understanding on X Search
Enable image understanding on X Search
Set enable_image_understanding on the x_search tool so Grok can inspect images attached to X posts when answering. Base X Search setup: Search X posts with the Grok API. For video on the same tool, see Enable video understanding on X Search.
xAI SDK
import os
from xai_sdk import Client
from xai_sdk.chat import user
from xai_sdk.tools import x_search
client = Client(api_key=os.getenv("XAI_API_KEY"))
chat = client.chat.create(
model="grok-4.6",
tools=[
x_search(enable_image_understanding=True),
],
)
chat.append(user("Find X posts with images about AI and describe what the images show."))
response = chat.sample()
print(response.content)
OpenAI Responses
curl https://api.x.ai/v1/responses \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $XAI_API_KEY" \
-d '{
"model": "grok-4.6",
"input": [
{
"role": "user",
"content": "Find X posts with images about AI and describe what the images show."
}
],
"tools": [
{
"type": "x_search",
"enable_image_understanding": true
}
]
}'
Vercel AI SDK
import { xai } from '@ai-sdk/xai';
import { generateText } from 'ai';
const { text } = await generateText({
model: xai.responses('grok-4.6'),
prompt: 'Find X posts with images about AI and describe what the images show.',
tools: {
x_search: xai.tools.xSearch({
enableImageUnderstanding: true,
}),
},
});
console.log(text);
Pitfalls
- This flag lives on
x_search. Web Search has its ownenable_image_understandingfor images found while browsing pages. - Enabling image understanding on Web Search also enables it for X Search when both tools are on the same request.
- Tool calls bill tokens plus invocations. One question can fire several searches.
- Console API credits are separate from SuperGrok's weekly pool on grok.com.