v2.2 · 2026-06-12
BGE-M3
bge-m3
BGE-M3 is a 0.567B embedding model profile for local AI planning. This page records provider, license, context, quantization, hardware-fit estimates, setup hints, source links, and caveats so readers do not choose a model by name alone.
PROVIDER
BAAI
FAMILY
BGE
MODEL TYPE
embedding
PARAMETERS
0.567B
MODALITIES
text
ARCHITECTURE
XLM-RoBERTa encoder
CONTEXT WINDOW
8,192 tokens
TRAINING TOKENS
—
RELEASE DATE
2024-02-05
GGUF REPOSITORIES
lm-kit/bge-m3-gguf community conversion
Community GGUF quantizations of BAAI/bge-m3; file table used for Q4_K_M, Q5_K_M, and Q8_0 sizes.
SETUP HINTS
OLLAMA
Use `ollama pull bge-m3`, then call Ollama's embedding endpoint with model `bge-m3`.LM STUDIO
Use an embedding workflow, not chat
LLAMA.CPP
Use an embedding-capable llama.cpp build/runtime and a vetted GGUF such as `lm-kit/bge-m3-gguf`QUANTIZATION FILE SIZES (GGUF)
Q4_K_M
0.438 GB
Q5_K_M
0.468 GB
Q8_0
0.635 GB
Quantization note: Q8_0 for quality; Q4_K_M for minimal local footprint
RAM / VRAM ESTIMATES
MIN RAM
2 GB
COMFORTABLE RAM
4 GB
These are conservative local-inference estimates, not official hardware requirements.
HARDWARE FIT
CPU only (no GPU)
good
8 GB RAM
good
16 GB RAM
good
32 GB RAM
good
8 GB VRAM
good
12 GB VRAM
good
24 GB VRAM
good
Apple Silicon (unified memory)
good
Hardware fit values are conservative local-inference estimates based on GGUF size plus runtime and KV-cache overhead. Actual requirements depend on context length, quantization, runtime, GPU offload, and other running apps.
BEST FOR
·Multilingual RAG and semantic search
AVOID IF
·You need a chat or code-generation model.
CAVEATS
·BGE-M3 is an embedding/retrieval model, not a chat or code-generation model.
·The releaseDate uses the public BGE-M3 paper publication date because the model card did not expose a separate exact model-release date.
SOURCE URLS
FIELD EVIDENCE
parameterSizeBollama.com/library/bge-m3
licensehuggingface.co/BAAI/bge-m3
contextWindowTokenshuggingface.co/BAAI/bge-m3
embeddingDimensionhuggingface.co/BAAI/bge-m3
embeddingModelhuggingface.co/BAAI/bge-m3
languageshuggingface.co/BAAI/bge-m3
q4FileSizeGbhuggingface.co/lm-kit/bge-m3-gguf
q5FileSizeGbhuggingface.co/lm-kit/bge-m3-gguf
q8FileSizeGbhuggingface.co/lm-kit/bge-m3-gguf
LANGUAGES
100+ working languages
CAPABILITIES
embedding model