📰 Multimodal News Reporter AI

  • Upload an audio recording and/or a relevant image; the AI will generate a news report you can revise and save.
  • Token output is set to 128 only for faster inference.
  • Note: This demo currently runs on CPU only.
  • Sample audio is trimmed to 10 seconds for faster inference.
  • Combined audio + image inference takes ~250-350 seconds; audio-only or image-only is much faster.
Click an example to test
Audio Interview Evidence Image Evidence