v2.2 · 2026-06-12

Qwen3 Embedding 4B

qwen3-embedding-4b

EMBEDDING

Qwen3 Embedding 4B is a 4B 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

Qwen / Alibaba

FAMILY

Qwen3 Embedding

MODEL TYPE

embedding

PARAMETERS

MODALITIES

text

ARCHITECTURE

—

CONTEXT WINDOW

32,768 tokens

TRAINING TOKENS

—

RELEASE DATE

2025-06-05

LICENSE & LINKS

Apache-2.0license text

MODEL CARD

huggingface.co/Qwen/Qwen3-Embedding-4B

HUGGING FACE

Qwen/Qwen3-Embedding-4B

GGUF REPOSITORIES

Qwen/Qwen3-Embedding-4B-GGUF official

Official GGUF repository referenced by the June 2026 model research.

SETUP HINTS

OLLAMA

ollama pull qwen3-embedding:4b

LLAMA.CPP

Use the cited GGUF repository (Qwen/Qwen3-Embedding-4B-GGUF) with current llama.cpp-compatible tooling.

QUANTIZATION FILE SIZES (GGUF)

Q4_K_M

2.5 GB

Q8_0

4.3 GB

Quantization note: Q4_K_M for first local attempts unless a cited runtime recommends a different default.

HARDWARE FIT

CPU only (no GPU)

8 GB RAM

unknown

16 GB RAM

unknown

32 GB RAM

unknown

8 GB VRAM

unknown

12 GB VRAM

unknown

24 GB VRAM

unknown

Apple Silicon (unified memory)

unknown

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

·Power users building multilingual knowledge bases who can spare extra disk, RAM, and indexing time.

AVOID IF

·Your vector database, hardware, or app expects small 768-dimensional embeddings and fast indexing on CPU.

CAVEATS

·This is not the first Qwen3 Embedding variant a beginner should try.

·The larger 2,560-dimensional vectors increase vector storage and indexing cost.

·Ollama reports 40K context for the 4B tag, but official Qwen specs list 32K; use 32K unless manually verified.

·Official GGUF lists Q5 variants, but Q5 file size was not confirmed in cited sources.

·Commercial-use status from the supplied research: Yes under Apache-2.0.

·Beginner summary from supplied research: Qwen3 Embedding 4B is the quality-oriented middle option for serious local RAG. It returns larger 2,560-dimensional vectors, handles long multilingual context, and fits advanced desktops better than casual beginner laptops or phones for local search.

SOURCE URLS