Back to changelog

Document Reading (PDFs & more)

Send PDFs and text-family documents to Gemini models via the OpenAI-compatible `file` content block.

LLM Gateway now supports document attachments on chat completions

LLM Gateway now accepts document attachments on chat completions through OpenAI's file content block. Send a PDF (or other supported file type) as base64 file_data and the gateway forwards it to the model.

1curl -X POST "https://api.llmgateway.io/v1/chat/completions" \2  -H "Authorization: Bearer $LLM_GATEWAY_API_KEY" \3  -H "Content-Type: application/json" \4  -d '{5    "model": "gemini-2.5-flash",6    "messages": [{7      "role": "user",8      "content": [9        { "type": "text", "text": "Summarize this document." },10        { "type": "file", "file": {11            "filename": "report.pdf",12            "file_data": "data:application/pdf;base64,JVBERi0xLjQK..."13        }}14      ]15    }]16  }'
  • Initial support on Google Gemini via Google AI Studio
  • New document capability flag on models — filter at /models?filters=1&document=true
  • Playground accepts file uploads and persists them across sessions and shares
  • Clean 400 errors when a model doesn't accept the MIME type, instead of opaque upstream failures

Read the docs →