> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getsmelt.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Supported Formats & Encoding

> File requirements, encoding support, and duplicate detection

## File Requirements

| Requirement         | Value           |
| ------------------- | --------------- |
| **Format**          | CSV only        |
| **Size limit**      | 50MB maximum    |
| **Maximum columns** | 100             |
| **Maximum rows**    | 50,000 per file |

## Supported Encodings

Smelt supports multiple character encodings:

| Encoding                 | Recommendation                     |
| ------------------------ | ---------------------------------- |
| **UTF-8**                | ✅ Recommended — best compatibility |
| **UTF-16**               | Supported                          |
| **Latin-1 (ISO-8859-1)** | Supported                          |
| **Windows-1252**         | Supported                          |

<Tip>
  When in doubt, use UTF-8. It handles international characters and is the most widely compatible.
</Tip>

## Exporting as UTF-8

<Tabs>
  <Tab title="From Excel">
    1. File → Save As
    2. Choose "CSV UTF-8 (Comma delimited)"
    3. Save
  </Tab>

  <Tab title="From Google Sheets">
    1. File → Download
    2. Choose "Comma Separated Values (.csv)"
    3. Google Sheets exports as UTF-8 by default
  </Tab>

  <Tab title="From Numbers (Mac)">
    1. File → Export To → CSV
    2. Text Encoding: Unicode (UTF-8)
    3. Export
  </Tab>
</Tabs>

## Duplicate Detection

Smelt automatically detects duplicates in your data:

### What's Detected

| Field             | Detection              |
| ----------------- | ---------------------- |
| **Emails**        | Exact match duplicates |
| **Company names** | Exact match duplicates |
| **Phone numbers** | Exact match duplicates |
| **LinkedIn URLs** | Exact match duplicates |

### Where Duplicates Are Shown

The file detail page displays:

* Duplicate count for each key field
* Which rows contain duplicates

### How Duplicates Affect Processing

<Note>
  Duplicates are still processed — each row generates output.
</Note>

Smelt doesn't automatically remove duplicates. Each row in your CSV will:

* Be processed by your templates
* Generate its own output
* Count toward your usage

<Tip>
  Consider deduplicating before upload if you want to:

  * Save on usage costs
  * Avoid sending duplicate emails
  * Keep your data clean
</Tip>

## Best Practices for CSV Formatting

<AccordionGroup>
  <Accordion title="Use headers in the first row">
    Row 1 should contain column names, not data.
  </Accordion>

  <Accordion title="Consistent column names">
    Keep naming consistent across files if you reuse templates.
  </Accordion>

  <Accordion title="No merged cells">
    CSVs don't support merged cells. Unmerge before exporting.
  </Accordion>

  <Accordion title="Clean data">
    Trim whitespace, fix obvious typos, remove empty rows.
  </Accordion>

  <Accordion title="Avoid special characters in headers">
    Simple alphanumeric column names work best.
  </Accordion>
</AccordionGroup>

## Common Encoding Issues

| Symptom                      | Cause                   | Fix                   |
| ---------------------------- | ----------------------- | --------------------- |
| Garbled characters (é, ’) | Wrong encoding detected | Re-save as UTF-8      |
| Missing characters           | Incompatible encoding   | Re-save as UTF-8      |
| Empty cells showing data     | Hidden characters       | Clean data, re-export |