alt.freeweights.embed / bge-m3-mlx-8bit

From:

mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.embedSubject:

bge-m3-mlx-8bit

Date: 2026-11-15Commit: 7eca4a1Signed: 2DC28CC7B26A283B ✓

Feature extraction model with 568M parameters. Safetensors weights, 621 MB in all. Published by mlx-community under MIT.

568M params

621MBbge-m3-mlx-8bitsafetensorspermissive

FactsParams568MSize621 MBFormatssafetensorsTaskfeature extractionLicenseMITFrom sourcepermissive

$ fw get mlx-community/bge-m3-mlx-8bit
# manifest sha256 61d853ad7733… · 621 MB · 11 files
$ fw verify ./bge-m3-mlx-8bit mlx-community/bge-m3-mlx-8bit

From sourceHosted (signed)mlx-community/bge-m3-mlx-8bitLast check: 2026-11-15 03:33 UTC

Run at home Get help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Fine-tune of bge-m3.

Works with

Run9 Fine-tune8 Convert8 Evaluate1 Scan2 Apps16

Runs with

Ranked by fit to this model (safetensors), then by GitHub stars.

Text Embeddings Inference

Transformers LocalAI tinygrad Triton Inference Server BentoML CTranslate2 ExLlamaV2 LitServe

Show all 16

sqlite-vec hnswlib pgvectorscale Infinity

Overview Model card Files Use it License Provenance Model tree

Overview

Size: 621 MB
Weights: 604 MB (safetensors)
Parameters: 568M
Formats: safetensors
Task: feature extraction
License: MIT (permissive)
Files: 11, each with a sha256
Publisher: mlx-community
Added: 2026-11-15
Section: alt.freeweights.embed

Lineage

Original release from the publisher.

Base model: BAAI/bge-m3

Where these bytes come from

Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources

Badge	Host	Last check	Result
From source	huggingface.co	2026-11-15 03:33 UTC	passed

Model card

From the publisher at commit 7eca4a1. Shown as plain text with links; images open at the publisher.

BGE-M3 MLX (8-bit Quantized)#

This is the BAAI/bge-m3 model converted to MLX format with 8-bit quantization for Apple Silicon.

Model Description#

BGE-M3 is a versatile embedding model capable of:

Dense retrieval
Sparse retrieval
Multi-vector (ColBERT) retrieval

This 8-bit quantized version offers the best quality among quantized variants.

Model Details#

Property	Value
Architecture	XLM-RoBERTa
Precision	8-bit (affine quantization)
Embedding Dimension	1024
Max Sequence Length	8192
Model Size	~592 MB
Quantization Group Size	64
Languages	100+ languages

Size Comparison#

Version	Size	Compression
FP16	1.1 GB	-
8-bit	592 MB	46%
6-bit	457 MB	58%
4-bit	321 MB	71%

Usage#

With MLX#

from mlx_embeddings.utils import load_model, load_tokenizer
import mlx.core as mx

model_path = "mlx-community/bge-m3-mlx-8bit"

# Load model and tokenizer
model = load_model(model_path)
tokenizer = load_tokenizer(model_path)

# Generate embeddings
text = "Hello, world!"
tokens = tokenizer.encode(text)
input_ids = mx.array([tokens])
output = model(input_ids)
embedding = output.last_hidden_state.mean(axis=1)  # Mean pooling

print(f"Embedding shape: {embedding.shape}")  # (1, 1024)

With oMLX#

curl http://127.0.0.1:8000/v1/embeddings \
  -H "Content-Type: application/json" \
  -d '{"model": "bge-m3-mlx-8bit", "input": "Your text here"}'

Quantization Details#

Method: Affine quantization
Bits per weight: 8
Group size: 64
Source: Converted from MLX FP16 version

Recommended Use Cases#

Quality-critical applications
Production deployments where quality is prioritized
Best choice when memory allows ~600MB

License#

MIT license (inherited from BAAI/bge-m3)

Citation#

@article{bge_m3,
  title={BGE M3-Embedding: Accurate, Efficient and Versatile Text Embedding},
  author={Chen, Jianlv and Xiao, Shitao and Zhang, Peitian and Luo, Kun and Zhang, Zheng},
  journal={arXiv preprint arXiv:2402.03216},
  year={2024}
}

Disclaimer#

This is an unofficial MLX conversion of the BAAI/bge-m3 model. For the original model, see BAAI/bge-m3.

Files

11 files · 621 MBWeights 604 MB (safetensors). Sizes and sha256 from the signed manifest.

Hide all 11 files with their sha256

File	Size	sha256
.gitattributes	1.57 KB	34448b82c17d60fec9b65b1f093c115ddbaadc04beb1b0140b6bfed2e012a930
README.md	2.65 KB	240e73c5bcac21a232a1d160ae438e80c1885afa6ab57a271f0093c50b6892a1
config.json	814 B	4603d4b7f4ed1c9aa3590c279d0558d426264463403847d293f66ade9f23dc1b
config_sentence_transformers.json	123 B	1eef72430e7194a1e59680e635aed81ffa083f05668dbc5bb1c56c04c0999c38
model.safetensors	604 MB	57b597e5aa8c102c2698cc0915760a839e4c1a30bff0a46d7883d27d3f010720
model.safetensors.index.json	51.2 KB	40240743a3a26832574855f41cf9a1babbd5b7ffa71cc462aa058be26f931883
modules.json	349 B	84e40c8e006c9b1d6c122e02cba9b02458120b5fb0c87b746c41e0207cf642cf
sentence_bert_config.json	54 B	eb9b44b13c0f52a3b3685c3b1cbdea1ba8b04bea123b98f61610048940776eb1
special_tokens_map.json	964 B	8c785abebea9ae3257b61681b4e6fd8365ceafde980c21970d001e834cf10835
tokenizer.json	17.1 MB	5df1f55d60c9705a501ab9a75550728625740741fe4be308dac4806c16b7d51d
tokenizer_config.json	379 B	c2ef99124628ae6f79a847ad67e5d3f5b016d0387de90c9fbaf17c45a98933af

VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.

Upstream commit	Date	Contents
7eca4a1c6e pinned now	2026-11-15 03:33 UTC	11 files · 621 MB

Use it

$ fw get mlx-community/bge-m3-mlx-8bit
# manifest sha256 61d853ad7733… · 621 MB · 11 files
$ fw verify ./bge-m3-mlx-8bit mlx-community/bge-m3-mlx-8bit

fw.pyChecks the signature and every sha256, tries each source in order, resumes.

$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/bge-m3-mlx-8bit.json --dir ./bge-m3-mlx-8bit
$ python3 fw.py verify ./bge-m3-mlx-8bit ./mlx-community__bge-m3-mlx-8bit.json

hfThe Hugging Face CLI with one setting. Same repo id.

$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/bge-m3-mlx-8bit --revision 7eca4a1c6ea1a0c5efc37598b369012f3985910f --local-dir ./bge-m3-mlx-8bit

transformersLoad the weights in Python.

import os
os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com"
from transformers import AutoModel, AutoTokenizer
tok = AutoTokenizer.from_pretrained("mlx-community/bge-m3-mlx-8bit", revision="7eca4a1c6ea1a0c5efc37598b369012f3985910f")
model = AutoModel.from_pretrained("mlx-community/bge-m3-mlx-8bit", revision="7eca4a1c6ea1a0c5efc37598b369012f3985910f", device_map="auto")

sentence-transformersEmbed text with sentence-transformers.

import os
os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com"
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("mlx-community/bge-m3-mlx-8bit", revision="7eca4a1c6ea1a0c5efc37598b369012f3985910f")
vectors = model.encode(["Hello"])

Run at home

License

Name: MIT License
Id: MIT
Class: Permissive
Redistribution: yes
Listing: Hosted (signed)
Decided: License as declared by the publisher. No person read it. 2026-11-15
Text in: README.md
Text sha256: 240e73c5bcac21a232a1d160ae438e80c1885afa6ab57a271f0093c50b6892a1
Terms gate: No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From:

mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.embedSubject: bge-m3-mlx-8bitDate: 2026-11-15Commit: 7eca4a1Signed: 2DC28CC7B26A283B ✓

Upstream: mlx-community/bge-m3-mlx-8bit
Pinned commit: 7eca4a1c6ea1a0c5efc37598b369012f3985910f
Manifest: bge-m3-mlx-8bit.json · .minisig
Manifest sha256: 61d853ad773370c4e151333881fb430b16e73adf50c10d42e6334daf6aebc016
Signature: minisign, Ed25519, key 2DC28CC7B26A283B
Created: 2026-11-15 03:33 UTC
Generator: freeweights-manifest/0.1.0
Takedowns: None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/mlx-community/bge-m3-mlx-8bit.json
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/bge-m3-mlx-8bit.json.minisig
$ minisign -Vm bge-m3-mlx-8bit.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq
$ shasum -a 256 <file>   # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

BAAI/bge-m3 · this model is a finetune of it

No model in the catalog names this one as its parent.

Get it

Download

$ HF_ENDPOINT=https://freeweights.allternit.com hf download mlx-community/bge-m3-mlx-8bit

Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest

minisign, key 2DC28CC7B26A283B. bge-m3-mlx-8bit.json. Provenance. Takedown history: none.

Source

huggingface.co/mlx-community/bge-m3-mlx-8bit

Related models

All of alt.freeweights.embed →

From: sentence-transformers <huggingface.co/sentence-transformers>Subject: all-MiniLM-L6-v1

From:

unsloth <huggingface.co/unsloth>Subject: all-MiniLM-L6-v2

From:

mlx-community <huggingface.co/mlx-community>Subject: all-MiniLM-L6-v2-4bit

From: sentence-transformers <huggingface.co/sentence-transformers>Subject: all-mpnet-base-v1

From: sentence-transformers <huggingface.co/sentence-transformers>Subject: allenai-specter

From: sentence-transformers <huggingface.co/sentence-transformers>Subject: bert-base-nli-max-tokens

bge-m3-mlx-8bit

Works with

Runs with

Fine-tune with

Convert with

Evaluate with

Scan with

Works with

Overview

Model card

BGE-M3 MLX (8-bit Quantized)#

Model Description#

Model Details#

Size Comparison#

Usage#

With MLX#

With oMLX#

Quantization Details#

Recommended Use Cases#

License#

Citation#

Disclaimer#

Files

Use it

License

Provenance

How to verify

Model tree

Related models