[MLsys2026 Best Paper]: https://arxiv.org/abs/2506.08276. RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate, and 100% private RAG application on your personal device.
Python aifaissgpt-osslangchainllama-indexllmlocalstorageoffline-firstollamaprivacypythonragretrieval-augmented-generationvector-databasevector-searchvectors
feat: smart embedding model defaults and incremental build fixes
yichuan520030910320956beb504d4e
diff --git a/packages/leann-core/src/leann/api.py b/packages/leann-core/src/leann/api.pyindex 9924f52..759e022 100644--- a/packages/leann-core/src/leann/api.py+++ b/packages/leann-core/src/leann/api.py@@ -357,7 +357,7 @@ class LeannBuilder:def __init__(self,backend_name: str,- embedding_model: str = "facebook/contriever",+ embedding_model: str = "sentence-transformers/all-MiniLM-L6-v2",dimensions: Optional[int] = None,embedding_mode: str = "sentence-transformers",embedding_options: Optional[dict[str, Any]] = None,diff --git a/packages/leann-core/src/leann/cli.py b/packages/leann-core/src/leann/cli.pyindex 3b6ed11..2f4b583 100644--- a/packages/leann-core/src/leann/cli.py+++ b/packages/leann-core/src/leann/cli.py@@ -29,6 +29,28 @@ from .settings import (from .sync import FileSynchronizer+def _default_embedding_model() -> str:+ """Pick a sensible default embedding model based on platform.++ | Platform | Default model |+ |------------|------------------------------------------------|+ | NVIDIA GPU | BAAI/bge-base-en-v1.5 |+ | macOS | sentence-transformers/all-MiniLM-L6-v2 |+ | Other/CPU | sentence-transformers/all-MiniLM-L6-v2 |+ """++ try:+ import torch++ if torch.cuda.is_available():+ return "BAAI/bge-base-en-v1.5"+ except ImportError:+ pass++ # macOS (MPS or CPU) and all other platforms: lightweight model+ return "sentence-transformers/all-MiniLM-L6-v2"++def _normalize_path(path: str) -> str:"""Return absolute path string for consistent keys."""if not path:@@ -209,11 +231,12 @@ Examples:choices=["hnsw", "diskann", "ivf"],help="Backend to use (default: hnsw)",)+ _default_model = _default_embedding_model()build_parser.add_argument("--embedding-model",type=str,- default="facebook/contriever",- help="Embedding model (default: facebook/contriever)",+ default=_default_model,+ help=f"Embedding model (default: {_default_model})",)build_parser.add_argument("--embedding-mode",@@ -2139,8 +2162,15 @@ Examples:is_compact = meta.get("is_compact", meta.get("backend_kwargs", {}).get("is_compact", True))++ # Normalize model names for comparison — e.g. "all-MiniLM-L6-v2"+ # should match "sentence-transformers/all-MiniLM-L6-v2"+ def _normalize_model_name(name: str) -> str:+ return name.rsplit("/", 1)[-1] if name else name+same_embedding = (- meta.get("embedding_model") == args.embedding_model+ _normalize_model_name(meta.get("embedding_model", ""))+ == _normalize_model_name(args.embedding_model)and meta.get("embedding_mode") == args.embedding_mode)@@ -2241,6 +2271,11 @@ Examples:else:self._log_rebuild_reason(meta, args, new_paths, removed_paths, modified_paths)+ # Force full reload — the partial all_texts from incremental+ # path only contains changed files, not the full corpus.+ all_texts = self.load_documents(+ docs_paths, args.file_types, include_hidden=args.include_hidden, args=args+ )# Full rebuild: load documents if not already loaded (first build or force)try:diff --git a/packages/leann-core/src/leann/mcp.py b/packages/leann-core/src/leann/mcp.pyindex dc1fe0d..c3d1e79 100755--- a/packages/leann-core/src/leann/mcp.py+++ b/packages/leann-core/src/leann/mcp.py@@ -208,6 +208,26 @@ def handle_build(request_id, args):if args.get("force", False):cmd.append("--force")+ # Auto-detect embedding model from existing index to ensure incremental+ # updates use the same model (avoids mismatch with CLI default).+ if index_name:+ import json as _json+ from pathlib import Path++ meta_path = Path.cwd() / ".leann" / "indexes" / index_name / "documents.leann.meta.json"+ if meta_path.exists():+ try:+ with open(meta_path, encoding="utf-8") as f:+ meta = _json.load(f)+ existing_model = meta.get("embedding_model")+ existing_mode = meta.get("embedding_mode")+ if existing_model:+ cmd.extend([f"--embedding-model={existing_model}"])+ if existing_mode:+ cmd.extend([f"--embedding-mode={existing_mode}"])+ except (OSError, ValueError):+ pass+rc, stdout, stderr = _run_leann(*cmd, timeout=600)if rc != 0: