majentik/Qwen3-Embedding-0.6B-MLX-4bit

Qwen3-Embedding-0.6B MLX 4-bit

MLX 4-bit quantization of Qwen/Qwen3-Embedding-0.6B, produced with mlx-embeddings on Apple Silicon.

What is this?

Qwen3-Embedding is a decoder-only LLM-style text embedding model from the Qwen3 family, using last-token pooling to produce dense vector representations. It scores near the top of MMTEB multilingual benchmarks while retaining Apache-2.0 licensing.

Quantization

• —Method: MLX affine quantization (mlx_embeddings.convert), group_size=64
• —Bits per weight: 4
• —Output size: 333 MB (vs ~1.2 GB for bf16 source)

Quickstart

from mlx_embeddings import load

model, tokenizer = load("majentik/Qwen3-Embedding-0.6B-MLX-4bit")

inputs = tokenizer(
    ["What is the capital of France?", "Paris is the capital of France."],
    padding=True, truncation=True, return_tensors="mlx"
)
outputs = model(inputs["input_ids"], attention_mask=inputs["attention_mask"])
embeddings = outputs.text_embeds  # already L2-normalised, shape [batch, dim]

For sentence similarity:

import mlx.core as mx

e = embeddings
scores = (e[0] @ e[1:].T).tolist()
print(scores)

Model Specifications

License

Apache 2.0 — inherited from the upstream Qwen3-Embedding model. Free for research and commercial use.

See also

• —Base: Qwen/Qwen3-Embedding-0.6B
• —Official GGUF: Qwen/Qwen3-Embedding-0.6B-GGUF (if published by Qwen)
• —mlx-embeddings package: https://github.com/Blaizzy/mlx-embeddings
• —Garden hub: majentik/garden
• —MTEB leaderboard: https://huggingface.co/spaces/mteb/leaderboard