Powerful Features for
Effortless Redirect Mapping
Every tool you need to handle site migrations, URL restructuring, and domain changes, powered by AI.
AI-Powered Semantic Matching
Our matching engine understands the meaning behind your URLs, not just the text. It works across languages, detects page types, and extracts key terms to find the right match even when URLs look completely different.
- Multilingual support: works with URLs in any language
- Entity extraction for colors, sizes, product types, and more
- Exact and fuzzy URL detection
/old-blog/seo-tips-2024
/insights/search-optimization
Different URLs, same semantic meaning, matched correctly.
5-Group Composite Scoring
Every URL pair is scored across five composite signal groups covering 14 sub-signals in total. Each group is weighted by importance, and the combined score gives you a clear confidence rating for every match. See the full signal reference in our documentation.
URL Structure 16%
- Page Type Match
- Path Structure
- Domain Match
URL Text 17%
- Semantic Embedding
- Fuzzy Match
Token Match 34%
- Slug Token Overlap
- Unique Term Match (Discriminative)
- Cross-Language Match
- Keyword Overlap
Content 18%
- Title Similarity
- H1 Similarity
- Meta Description
Identity 15%
- Numeric ID Match
- Locale Match
Confidence Classification
Every match is classified into a confidence tier so you know which redirects are ready to implement and which need manual review. Learn more about score thresholds and export options.
*Actual distribution depends on how similar your source and destination URLs are.
Built for Scale
Handle thousands of URLs without breaking a sweat. Batch processing, optimized top-K selection, and efficient embedding pipelines keep everything fast.
- Batch embedding for 1000+ URLs
- Real-time progress tracking
- Summary & detailed CSV export
*Processing times vary based on dataset size, available hardware, and chosen model.
Frequently Asked Questions
What is redirect mapping?
Redirect mapping is the process of matching old URLs to new URLs during a site migration, CMS change, or domain move. Each old URL gets a 301 redirect pointing to the most relevant new page, preserving search rankings and user experience.
How does the AI matching work?
Redirecta uses a multilingual sentence-transformer model to understand the semantic meaning of each URL and its metadata (titles, headings, meta descriptions). Every source-destination pair is scored across 5 signal groups covering 14 sub-signals: URL structure, URL text, token matching, content similarity, and identity. The highest-scoring destination is selected as the best match.
What file format do I need?
A CSV file with at least two columns: source (old URLs) and
destination (new URLs). You can optionally include page titles, H1 headings,
and meta descriptions for better accuracy. See the
CSV format guide for details.
Does it work with non-English URLs?
Yes. The default multilingual model supports 50+ languages, and 12 built-in dictionaries handle cross-language entity matching for colors, sizes, materials, page types, clothing, furniture, electronics, food and drink, home, and more. It works out of the box for migrations between localized sites.
How many URLs can I process?
The tool is optimized for large datasets. You can process thousands of URLs per job. Processing time depends on dataset size and your chosen thoroughness setting, but a typical job with 1,000-5,000 URLs completes in a few minutes.
Is it free to use?
Yes. Redirecta is free to use with no credit card required. Create an account, upload your CSV, and start mapping. You can run up to 10 jobs at a time.
Ready to Try It?
Upload your first CSV and see the results in minutes. No credit card required.