Document Preview
Invoice preview will render here
Waiting for document...
Extracted Data
Structured output parsed natively from visual layout.
Confidence: —
Line Items
| Description | Qty | Rate | Amount |
|---|---|---|---|
| No items extracted yet. | |||
Multimodal Vision
Utilizes Llama 3.2 Vision, a state-of-the-art multimodal model that understands spatial
document layout, logos, and dense tables natively.
Complete Privacy
Processing is fully local. Zero data leaves your infrastructure, making it perfect for
sensitive financial data.
Local Extensibility
Easily swap or fine-tune other local vision models (like Llava or Qwen2-VL) as your
requirements evolve.