Files

Manage binary files (PDFs, CSVs, images, documents, etc.) attached to your Mentiora project. Files are stored server-side and can be used for analysis, knowledge bases, or any custom workflow.

Upload a File

TypeScript

import { readFileSync } from 'fs';

const result = await client.files.upload({
  filename: 'report.pdf',
  content: readFileSync('./report.pdf'),
  mimeType: 'application/pdf',
  description: 'Q1 sales report',
});

console.log(`Uploaded: ${result.fileId}`);

Python

Async (recommended):

from mentiora import UploadFileParams

with open('report.pdf', 'rb') as f:
    result = await client.files.upload_async(UploadFileParams(
        filename='report.pdf',
        content=f.read(),
        mime_type='application/pdf',
        description='Q1 sales report',
    ))

print(f'Uploaded: {result.file_id}')

Sync:

from mentiora import UploadFileParams

with open('report.pdf', 'rb') as f:
    result = client.files.upload(UploadFileParams(
        filename='report.pdf',
        content=f.read(),
        mime_type='application/pdf',
        description='Q1 sales report',
    ))

print(f'Uploaded: {result.file_id}')

The SDK handles base64 encoding automatically. Maximum file size is 50 MB.

List Files

TypeScript

// List all files
const files = await client.files.list();
console.log(`Total: ${files.totalCount}`);
for (const file of files.data) {
  console.log(`${file.filename} (${file.size} bytes)`);
}

// Search by name or description
const results = await client.files.list({ query: 'report' });

// Paginate
const page2 = await client.files.list({ offset: 10, count: 10 });

Python

Async (recommended):

# List all files
files = await client.files.list_async()
print(f'Total: {files.total_count}')
for file in files.data:
    print(f'{file.filename} ({file.size} bytes)')

# Search by name or description
results = await client.files.list_async(query='report')

# Paginate
page2 = await client.files.list_async(offset=10, count=10)

Sync:

files = client.files.list()
results = client.files.list(query='report')
page2 = client.files.list(offset=10, count=10)

Get File Metadata

TypeScript

const file = await client.files.get('019d6d5d-d384-7793-9e0f-2592a9986021');
console.log(`${file.filename} — ${file.mimeType} — ${file.size} bytes`);

Python

Async (recommended):

file = await client.files.get_async('019d6d5d-d384-7793-9e0f-2592a9986021')
print(f'{file.filename} — {file.mime_type} — {file.size} bytes')

Sync:

file = client.files.get('019d6d5d-d384-7793-9e0f-2592a9986021')

Download a File

TypeScript

const content = await client.files.download('019d6d5d-d384-7793-9e0f-2592a9986021');
// content is a Uint8Array — write to disk or process in memory
import { writeFileSync } from 'fs';
writeFileSync('./downloaded-report.pdf', content);

Python

Async (recommended):

content = await client.files.download_async('019d6d5d-d384-7793-9e0f-2592a9986021')
# content is bytes — write to disk or process in memory
with open('downloaded-report.pdf', 'wb') as f:
    f.write(content)

Sync:

content = client.files.download('019d6d5d-d384-7793-9e0f-2592a9986021')
with open('downloaded-report.pdf', 'wb') as f:
    f.write(content)

Delete a File

Deletion fails with a 409 Conflict error if the file is still referenced by a knowledge base document. Remove the reference first, then delete.

TypeScript

const result = await client.files.delete('019d6d5d-d384-7793-9e0f-2592a9986021');
console.log(`Deleted: ${result.deleted}`);

Python

Async (recommended):

result = await client.files.delete_async('019d6d5d-d384-7793-9e0f-2592a9986021')
print(f'Deleted: {result.deleted}')

Sync:

result = client.files.delete('019d6d5d-d384-7793-9e0f-2592a9986021')

Supported File Types

The platform accepts a wide range of file types including:

• Documents: PDF, DOCX, DOC, ODT, RTF
• Structured data: CSV, TSV, JSON, JSONL, XML, YAML
• Text: TXT, Markdown, HTML
• Images: PNG, JPEG, GIF, WebP, SVG, TIFF, BMP
• Email: EML
• eBooks: EPUB

See Also

• Files API Reference -- full method and type reference
• Authentication -- setting up your API key
• Examples -- complete, runnable code samples