Find Every Difference Between
Two CSV Files. Instantly.
Primary key matching finds added, deleted, and modified rows — even when files are in different order. Cell-level precision. Zero uploads. Works on files that crash Excel.
You've done this the hard way before
VLOOKUP Takes 20+ Minutes
Open both files. Write VLOOKUP formulas. Add a change flag. Filter mismatches. Repeat for every column you care about. Then Excel crashes on row 1,048,577.
Manual Review Misses Things
Scrolling row by row, eyeballing values, copy-pasting into a separate sheet. A single transposed digit slips through. The audit fails.
Online Diff Tools Upload Your Files
Diffchecker and other web tools require an upload. Customer records. Financial exports. Healthcare data. You're not supposed to do that.
Everything VLOOKUP Can't Do
Structured CSV comparison, not text matching. Why VLOOKUP breaks at scale →
Primary Key Matching
Match rows by a unique identifier (ID, Email, SKU). Finds added, deleted, and modified rows correctly even when files are sorted differently.
Cell-Level Change Detection
For each modified row, see exactly which cells changed with old and new values side-by-side. No column-by-column scanning required.
Ignore Timestamp Columns
Exclude created_at, updated_at, or any audit field from comparison. Compare what changed in the data — not when the row was touched.
Export Diff Report
Download the full comparison as CSV (for Excel or database import) or JSON (for developer pipelines). Labeled: added, deleted, modified.
Handles Files Excel Can't Open
Streaming architecture processes 5M+ rows without loading the full dataset into memory. Files over Excel's 1M row limit compare normally.
Zero Upload Privacy
Both files are processed entirely in your browser. Nothing is transmitted to any server. Safe for healthcare, financial, and customer data.
SplitForge vs Every Alternative
Compared to Excel VLOOKUP, Python pandas, and Diffchecker
| Feature | SplitForge | Excel VLOOKUP | Python pandas | Diffchecker |
|---|---|---|---|---|
| Max file size | 1GB+ (10M+ rows) | ~50MB before crashes | Unlimited (RAM-bound) | Strict limits (varies by tier) |
| Setup required | None — open browser | Excel installed | Python + pandas + code | None |
| Files uploaded to server | Never | Never | Never | Typically yes — verify before use |
| Primary key matching | Yes — built in | Manual VLOOKUP setup | Code: df.merge() | No |
| Cell-level change detection | Yes — automatic | Manual per column | Code required | Text diff only |
| Ignore specific columns | Yes (comma list) | Manual delete columns | Code: drop() columns | No |
| Export diff report | CSV or JSON | Manual copy/paste | Code: to_csv() | Text copy only |
| Speed (1M rows) | 4.8 sec | 10+ minutes | ~30 seconds | N/A at this scale |
| Row order flexibility | Yes (primary key mode) | Must sort first | Yes (merge key) | No (line-by-line only) |
| Cost | Free | Paid (Microsoft 365 subscription) | Free (technical barrier) | Free (limited) / paid tiers |
Who This Is (And Isn't) For
Use SplitForge If…
- Your files contain sensitive data that cannot be uploaded anywhere
- You compare database exports, CRM snapshots, or audit logs
- Your files exceed Excel's 1M row limit or cause crashes
- You need cell-level precision — not just "row 42 changed"
- You compare CSV files regularly and want it done in 30 seconds
- You need a structured diff report (CSV/JSON) for documentation or import
Use Something Else If…
- You need to compare JSON, XML, or database tables directly→ use a database diff tool or Python jq
- You need automated, scheduled comparisons in a pipeline→ use Python pandas, dbt, or Apache Spark
- Your files are 50M+ rows (requires 10GB+ browser RAM)→ use cloud ETL or database JOINs
- You want version control for CSV data over time→ use DVC, LakeFS, or git-csv-diff
- You need to merge the diff back into a single file→ use Python pandas merge with conflict resolution
Our position on data privacy
The CSV tooling industry normalized uploading sensitive data to third-party servers. Healthcare records. Financial exports. Customer lists. All uploaded by default. We reject that assumption. SplitForge processes everything client-side — not as a feature, but as the ethical default. Why we built it this way →
How Much Time Are You Losing?
Baseline: ~20 min per manual Excel VLOOKUP comparison session
Estimate based on 20-min manual VLOOKUP workflow vs ~30-sec SplitForge end-to-end time (file upload + comparison + export). Individual results vary.
Built for Real Comparison Workflows
Data Pipeline Auditing
Compare a source export to a downstream export to verify a pipeline didn't drop, duplicate, or corrupt records between stages.
CRM Snapshot Diff
Compare this week's CRM export to last week's. Find new contacts, deleted records, and changed fields — without VLOOKUP formulas.
Finance Reconciliation
Compare two ledger exports to find discrepancies before month-end close. Files stay on your machine — no upload to any SaaS tool.
Healthcare Record Auditing
Compare patient record exports across time periods. Files stay in your browser — no upload to any server, supporting data handling requirements for sensitive records.
Product Catalog Updates
Compare a vendor's updated catalog CSV to your existing one. Find new SKUs, price changes, and discontinued products in seconds.
Database Migration Validation
Export data before and after a migration. Compare to verify every row migrated correctly before decommissioning the old system.
Edge Cases, Handled
Real-world CSV files are messy. Here's how SplitForge handles the tricky ones.
How Teams Use CSV Compare
Illustrative workflows based on common use cases — not attributed quotes.
Weekly CRM Reconciliation
Writing VLOOKUP formulas in Excel every Monday to find contacts that changed, were added, or went inactive since the previous week's export.
Drop last week's and this week's export. Pick the email column as primary key. Done in under a minute — diff exported as CSV for the team.
Post-Migration Record Validation
After a database migration, spot-checking a sample of rows manually. No way to verify all 2M+ records migrated correctly without a dedicated diff tool.
Export before-and-after as CSVs. Run primary key comparison against the record ID column. Full diff report shows exactly which rows changed or were lost.
Monthly Vendor Catalog Update
Vendor sends a new catalog CSV each month. Finding price changes, new SKUs, and discontinued items requires manual VLOOKUP or upload to a web tool.
Compare current and new catalog CSVs. Filter modified rows to find price changes. Filter added rows to find new SKUs. Catalog data never leaves the machine.
Perfect For
- Data analysts doing weekly or monthly audit checks
- Finance teams reconciling exports before period close
- Healthcare teams auditing patient record exports (sensitive data stays local)
- Data engineers validating pipeline output vs source
- Operations teams tracking CRM or ERP snapshot changes
- E-commerce teams diffing vendor catalog updates
Honest Limitations
- 10M+ row comparisons require 2–3GB browser RAM — may fail on low-memory machines
- Line-by-line mode requires files to be in identical sort order
- CSV and TSV only — not native Excel XLSX (convert first using our Excel to CSV Converter)
- No persistent history — comparison results are not saved between sessions
- No scheduled or automated runs — manual browser tool only
- For 50M+ row files or CI/CD pipelines, use Python pandas or a database JOIN
Frequently Asked Questions
Stop Fighting VLOOKUP. Just Compare.
Drop your two CSV files. Pick your primary key. Click Compare. Get a complete diff report in seconds — with your data never leaving your machine.