Paste the example CSV below, select comma as the delimiter and enable the header row. Convert it to an array of objects. With the initial string mappings, name becomes a key and age remains a string, as the JSON example shows. Compare the output with the source record before changing any type or nesting settings.
Use column mapping when a field needs another name or a different type. Keep identifiers with leading zeros as strings. For nested objects, enable nesting for the mapped field and use the configured path separator; a dot in a header alone is not a complete conversion policy. Convert again after changing mappings, then inspect empty cells, numbers and date-like values before downloading the JSON for another application.
Keep your mappings with the source so another run uses the same field names and intended output types.
name,email,age
Alex,[email protected],30
Jordan,[email protected],25
[{"name":"Alex","email":"[email protected]","age":"30"},{"name":"Jordan","email":"[email protected]","age":"25"}]
Object output uses headers as property names and creates an object per row. Without headers, names are generated. Array output retains rows as arrays. Choose the receiving program’s required structure. Duplicate or empty headers cannot reliably name distinct object properties; review them first. Valid JSON alone does not establish compatibility with an API schema.
Column mappings can rename fields, choose a type and mark a mapped path as nested. Number conversion keeps values that do not match the supported numeric form as strings. Boolean and date conversions have their own interpretation rules. Inspect representative values instead of assuming every text cell will become the intended type.
With string mappings, nonempty cells remain text. Explicit conversion of an empty or whitespace-only cell can produce null. Quoted separators and multiline values are parsed as CSV fields rather than split blindly at each comma or newline. Keep postal codes, phone numbers and account identifiers as strings when their spelling matters.
The file upload allows 100 MiB. Parsed rows and JSON output occupy browser memory, so complexity matters alongside size. The converter does not stream arbitrary datasets to an API or create database records. Test a sample, then inspect first, last, and unusual records before using the complete output.
Processing runs in your browser. Choosing Share uploads the selected tool state; anyone with the link can open it. Signing in to save sends content to your account library and records save history. Review the selection before either action. Closing the page does not delete those copies. Copying locally does not create a server share.
These are output configurations, not different CSV input types. All work with standard CSV files but optimize the JSON output for different use cases:
General-purpose conversion with pretty-printed JSON for readability. Best for most users.
Minified JSON output without extra spacing. Optimized for REST APIs and data transmission.
Emphasizes hierarchical structure. Use when your CSV has dot notation columns (e.g., "user.name").
Outputs arrays of arrays instead of objects. Maintains tabular structure, useful for data processing.
Upload a file, paste your data, or try sample data with standard conversion format