AI-powered code search that actually understands your codebase structure. Find functions by meaning, not keywords.
grep and ripgrep only match text. They don't understand code structure or semantics.
$ grep -r "validate" .
Returns:
verify_token(), check_input() (semantic matches)Query: "Where do we validate user input?"
Returns (ranked by relevance):
validate_user_input() [src/auth.rs:142]User::validate() [src/models.rs:89]impl Validate for LoginForm [src/forms.rs:34]sanitize_input() [src/security.rs:67] (related)Automatically skips node_modules, target/, .git, and respects your .gitignore. Only indexes what matters.
Tree-sitter extracts functions, classes, methods with full context. Smart boundaries keep code together.
CodeBERT-style model processes 128 chunks at once on GPU/CPU. 40x faster than sequential.
HNSW for semantic similarity + Tantivy BM25 for keywords. Fusion ranking combines both.
BLAKE3 hashing detects changed files. Only reindex what's new. Full codebase updates in seconds.
<200MB index for typical 50K file project. Streaming pipeline keeps memory usage constant.
| Metric | ripgrep | H²CE-v2 |
|---|---|---|
| Search Latency | 15ms | 35ms |
| Semantic Understanding | ❌ No | ✅ Yes |
| MRR@10 (Accuracy) | 0.31 | 0.74 |
| Index 100K files | N/A | 4m 22s |
| Memory Usage | ~50MB | 184MB |
Transform how you search code. 30-day money-back guarantee.
one-time payment
Or $3/month subscription