Qwen3 Embedding 4B GGUF

🚀 Qwen3-Embedding-4B GGUF Models

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It offers comprehensive text embeddings and reranking models in various sizes, inheriting the excellent multilingual capabilities, long - text understanding, and reasoning skills of its foundational model.

🚀 Quick Start

Prerequisites

With Transformers versions earlier than 4.51.0, you may encounter the following error:

KeyError: 'qwen3'

Installation

Ensure you have transformers>=4.51.0 installed.

Usage

There are two main ways to use the model:

Sentence Transformers Usage

# Requires transformers>=4.51.0

from sentence_transformers import SentenceTransformer

# Load the model
model = SentenceTransformer("Qwen/Qwen3-Embedding-4B")

# We recommend enabling flash_attention_2 for better acceleration and memory saving,
# together with setting `padding_side` to "left":
# model = SentenceTransformer(
#     "Qwen/Qwen3-Embedding-4B",
#     model_kwargs={"attn_implementation": "flash_attention_2", "device_map": "auto"},
#     tokenizer_kwargs={"padding_side": "left"},
# )

# The queries and documents to embed
queries = [
    "What is the capital of China?",
    "Explain gravity",
]
documents = [
    "The capital of China is Beijing.",
    "Gravity is a force that attracts two bodies towards each other. It gives weight to physical objects and is responsible for the movement of planets around the sun.",
]

# Encode the queries and documents. Note that queries benefit from using a prompt
# Here we use the prompt called "query" stored under `model.prompts`, but you can
# also pass your own prompt via the `prompt` argument
query_embeddings = model.encode(queries, prompt_name="query")
document_embeddings = model.encode(documents)

# Compute the (cosine) similarity between the query and document embeddings
similarity = model.similarity(query_embeddings, document_embeddings)
print(similarity)
# tensor([[0.7534, 0.1147],
#         [0.0320, 0.6258]])

Transformers Usage

# Requires transformers>=4.51.0
import torch
import torch.nn.functional as F

from torch import Tensor
from transformers import AutoTokenizer, AutoModel


def last_token_pool(last_hidden_states: Tensor,
                 attention_mask: Tensor) -> Tensor:
    left_padding = (attention_mask[:, -1].sum() == attention_mask.shape[0])
    if left_padding:
        return last_hidden_states[:, -1]
    else:
        sequence_lengths = attention_mask.sum(dim=1) - 1
        batch_size = last_hidden_states.shape[0]
        return last_hidden_states[torch.arange(batch_size, device=last_hidden_states.device), sequence_lengths]


def get_detailed_instruct(task_description: str, query: str) -> str:
    return f'Instruct: {task_description}\nQuery:{query}'

# Each query must come with a one-sentence instruction that describes the task
task = 'Given a web search query, retrieve relevant passages that answer the query'

queries = [
    get_detailed_instruct(task, 'What is the capital of China?'),
    get_detailed_instruct(task, 'Explain gravity')
]
# No need to add instruction for retrieval documents
documents = [
    "The capital of China is Beijing.",
    "Gravity is a force that attracts two bodies towards each other. It gives weight to physical objects and is responsible for the movement of planets around the sun."
]
input_texts = queries + documents

tokenizer = AutoTokenizer.from_pretrained('Qwen/Qwen3-Embedding-4B', padding_side='left')
model = AutoModel.from_pretrained('Qwen/Qwen3-Embedding-4B')

# We recommend enabling flash_attention_2 for better acceleration and memory saving.
# model = AutoModel.from_pretrained('Qwen/Qwen3-Embedding-4B', attn_implementation="flash_attention_2", torch_dtype=torch.float16).cuda()

max_length = 8192

# Tokenize the input texts
batch_dict = tokenizer(
    input_texts,
    padding=True,
    truncation=True,
    max_length=max_length,
    return_tensors="pt",
)
batch_dict.to(model.device)
outputs = model(**batch_dict)
embeddings = last_token_pool(outputs.last_hidden_state, batch_dict['attention_mask'])

# normalize embeddings
embeddings = F.normalize(embeddings, p=2, dim=1)
scores = (embeddings[:2] @ embeddings[2:].T)
print(scores.tolist())
# [[0.7534257769584656, 0.1146894246339798], [0.03198453038930893, 0.6258305311203003]]

💡 Usage Tip

We recommend that developers customize the instruct according to their specific scenarios, tasks, and languages. Our tests have shown that in most retrieval scenarios, not using an instruct on the query side can lead to a drop in retrieval performance by approximately 1% to 5%.

✨ Features

Model Generation Details

This model was generated using llama.cpp at commit 1f63e75f.

Quantization beyond the IMatrix

Testing a new quantization method using rules to bump important layers above what the standard imatrix would use. The standard IMatrix does not perform very well at low bit quantization and for MOE models. So, llama.cpp --tensor-type is used to bump up selected layers. See Layer bumping with llama.cpp. This creates larger model files but increases precision for a given model size.

Choosing the Right Model Format

Selecting the correct model format depends on your hardware capabilities and memory constraints.

Qwen3 Embedding Series Highlights

• Exceptional Versatility: The embedding model has achieved state-of-the-art performance across a wide range of downstream application evaluations. The 8B size embedding model ranks No.1 in the MTEB multilingual leaderboard (as of June 5, 2025, score 70.58), while the reranking model excels in various text retrieval scenarios.
• Comprehensive Flexibility: The Qwen3 Embedding series offers a full spectrum of sizes (from 0.6B to 8B) for both embedding and reranking models, catering to diverse use cases that prioritize efficiency and effectiveness. Developers can seamlessly combine these two modules. Additionally, the embedding model allows for flexible vector definitions across all dimensions, and both embedding and reranking models support user-defined instructions to enhance performance for specific tasks, languages, or scenarios.
• Multilingual Capability: The Qwen3 Embedding series offer support for over 100 languages, thanks to the multilingual capabilities of Qwen3 models. This includes various programming languages, and provides robust multilingual, cross-lingual, and code retrieval capabilities.

📚 Documentation

Model Overview

Qwen3-Embedding-4B has the following features:

For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub.

Qwen3 Embedding Series Model list

Note:

• MRL Support indicates whether the embedding model supports custom dimensions for the final embedding.
• Instruction Aware notes whether the embedding or reranking model supports customizing the input instruction according to different tasks.
• Our evaluation indicates that, for most downstream tasks, using instructions (instruct) typically yields an improvement of 1% to 5% compared to not using them. Therefore, we recommend that developers create tailored instructions specific to their tasks and scenarios. In multilingual contexts, we also advise users to write their instructions in English, as most instructions utilized during the model training process were originally written in English.

🔧 Technical Details

Evaluation - MTEB (Multilingual)

Note: For compared models, the scores are retrieved from MTEB online leaderboard

📄 License

This model is licensed under the apache-2.0 license.