Document Parsing Service Guide

Our document parsing service leverages the claude-3-7-sonnet-latest,deepseek-r1,mistral-large-latest model and XAI XAPI to extract and analyze content from various document formats using AI models (including Anthropic Claude series). The service can process single or multiple documents and provide relevant information based on user queries.

Supported File Formats

We support parsing for the following 7 document formats:

PDF (.pdf)
Text files (.txt, .sh, .py, .js, etc.)
Word documents (.doc, .docx)
Excel spreadsheets (.xls, .xlsx)
CSV files (.csv)
Markdown files (.md, .markdown)
HTML files (.html, .htm)

Usage Limits

PDF file size must not exceed 4.5MB
Maximum of 5 files per conversation

API Endpoints

Our service provides two main API endpoints:

${BASE_URL}/v1/messages
${BASE_URL}/v1/chat/completions

Recommended setup:

export BASE_URL="https://api.xairouter.com"      # SaaS single-tenant
# export BASE_URL="https://api.xaicontrol.com"   # BYOK multi-tenant

Usage Examples

Here are several examples of using our document parsing service:

Example 1: Single Document Processing

Using the /v1/messages endpoint to process a single document:

API_KEY="YOUR_XAI_API_KEY"

curl ${BASE_URL}/v1/messages \
   -H "content-type: application/json" \
   -H "x-api-key: $API_KEY" \
   -H "anthropic-version: 2023-06-01" \
   -d '{
     "model": "claude-3-7-sonnet-latest",
     "max_tokens": 1024,
     "messages": [{
         "role": "user",
         "content": [{
             "type": "document",
             "source": {
                 "type": "url",
                 "url": "https://cfbed.1314883.xyz/file/1744972227336_test-1.pdf"
             }
         },
         {
             "type": "text",
             "text": "Extract key information from the document"
         }]
     }]
 }'

Example 2: Single Document Processing with Alternative Endpoint

Using the /v1/chat/completions endpoint to process a single document:

API_KEY="YOUR_XAI_API_KEY"
curl -X POST "${BASE_URL}/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $API_KEY" \
  -d '{
    "model": "claude-3-7-sonnet-latest",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "document",
            "source": {
              "type": "url",
              "url": "https://cfbed.1314883.xyz/file/1744972223519_test.docx"
            }
          },
          {
            "type": "text",
             "text": "Summarize the main content of this document"
          }
        ]
      }
    ],
    "max_tokens": 1000
  }'

Example 3: Multi-Document Processing

Using the /v1/chat/completions endpoint to process multiple documents (max 5):

API_KEY="YOUR_XAI_API_KEY"
curl -X POST "${BASE_URL}/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $API_KEY" \
  -d '{
    "model": "claude-3-7-sonnet-latest",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "document",
            "source": {
              "type": "url",
              "url": "https://cfbed.1314883.xyz/file/1744972223519_test.xlsx"
            }
          },
          {
            "type": "document",
            "source": {
              "type": "url",
              "url": "https://cfbed.1314883.xyz/file/1744972227336_test.pdf"
            }
          },
          {
            "type": "text",
             "text": "Compare the data differences between these documents"
          }
        ]
      }
    ],
    "max_tokens": 1000
  }'

Example 4: Multi-Document Processing with Alternative Endpoint

Using the /v1/messages endpoint to process multiple documents:

API_KEY="YOUR_XAI_API_KEY"
curl ${BASE_URL}/v1/messages \
   -H "content-type: application/json" \
   -H "x-api-key: $API_KEY" \
   -H "anthropic-version: 2023-06-01" \
   -d '{
     "model": "claude-3-7-sonnet-latest",
     "max_tokens": 1024,
     "messages": [{
         "role": "user",
         "content": [{
             "type": "document",
             "source": {
                 "type": "url",
                 "url": "https://cfbed.1314883.xyz/file/1744972227336_test.md"
             }
         },
         {
             "type": "document",
             "source": {
                 "type": "url",
                 "url": "https://cfbed.1314883.xyz/file/1744972223519_test.html"
             }
         },
         {
             "type": "text",
             "text": "Analyze the content structure of these documents"
         }]
     }]
 }'

Note: Ensure all documents comply with size limits and do not exceed 5 files per request.