pdf-summarizer
v0.9.0Returns a structured summary, key points, and extracted entities from any PDF document. Handles multi-page documents up to 200 pages.
Install
bloom install pdf-summarizer@0.9.0
Stars2
Downloads2751
Version0.9.0
PublishedMar 4, 2026
README
Overview
Pass any PDF URL or base64-encoded document to pdf-summarizer and receive a concise summary, bullet-point key findings, and extracted named entities.
Usage
{
"url": "https://arxiv.org/pdf/2303.08774.pdf",
"maxLength": 500,
"extractEntities": true
}
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| url | string | yes* | Public URL to the PDF |
| base64 | string | yes* | Base64-encoded PDF (alternative to url) |
| maxLength | number | no | Max summary word count (default: 300) |
| extractEntities | boolean | no | Extract names, orgs, dates (default: false) |
| language | string | no | Target language for output (default: "en") |
*One of url or base64 is required.
License
MIT
Agent Schema
{
"name": "pdf-summarizer",
"inputs": {
"type": "object",
"properties": {
"url": {
"type": "string",
"format": "uri"
},
"base64": {
"type": "string"
},
"maxLength": {
"type": "number",
"default": 300
},
"extractEntities": {
"type": "boolean",
"default": false
}
}
},
"outputs": {
"type": "object",
"properties": {
"summary": {
"type": "string"
},
"entities": {
"type": "object"
},
"keyPoints": {
"type": "array",
"items": {
"type": "string"
}
},
"pageCount": {
"type": "number"
},
"wordCount": {
"type": "number"
}
}
},
"runtime": "python",
"version": "0.9.0",
"description": "Summarizes PDF documents",
"capabilities": [
"pdf",
"summarization",
"nlp"
]
}