Skip to content
Weights Not Walls
Menu

Qwen

Qwen3-4B-Base

Reviewed
Size
4.0B parameterscounted from the safetensors header on the Hub
Licence
unknown, Apache License 2.0Our reading, check the licence: You can use, change and redistribute the model, including commercially. Keep the licence and any notices, and say which files you changed; you do not have to publish your changes. It includes a patent licence that ends for anyone who sues over patents in the work.
On the Hub since
28 Apr 2025the date the repo was created
Context length
32k tokensfrom the model's config
Kind
Base model (no base model declared on the Hub)Qwen3ForCausalLM
On Hugging Face
Qwen/Qwen3-4B-Base414k downloads in the last 30 days, 119 likes

Scores

No independent scores for this model in Epoch AI's data. A score is only shown here when Epoch's name for a model matches this repo with certainty; an uncertain match is left out rather than guessed.

Claimed by the maker Claimed

The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.

Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.

Run it yourself

file sizes read from the Hub 19 Nov 2026

Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 1.1 GB, worked out from the model's config (Qwen/Qwen3-4B-Base), at 16-bit precision. Choose a machine above to see a verdict for each format. Method

GGUF for llama.cpp, Ollama, LM Studio and others

mradermacher/Qwen3-4B-Base-GGUF (converted by mradermacher), 4.4k downloads. Sizes are the real files on the Hub, all parts added up.

FormatSizeMemory neededGet it
Q2_K1.6 GBfile size3.2 GB1.6 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q2_K
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q2_K
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q2_K.gguf" --local-dir Qwen3-4B-Base-GGUF
Q3_K_S1.8 GBfile size3.4 GB1.8 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q3_K_S
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q3_K_S
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q3_K_S.gguf" --local-dir Qwen3-4B-Base-GGUF
Q3_K_M1.9 GBfile size3.6 GB1.9 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q3_K_M
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q3_K_M
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q3_K_M.gguf" --local-dir Qwen3-4B-Base-GGUF
Q3_K_L2.1 GBfile size3.7 GB2.1 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q3_K_L
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q3_K_L
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q3_K_L.gguf" --local-dir Qwen3-4B-Base-GGUF
IQ4_XS2.1 GBfile size3.8 GB2.1 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:IQ4_XS
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:IQ4_XS
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.IQ4_XS.gguf" --local-dir Qwen3-4B-Base-GGUF
Q4_K_S2.2 GBfile size3.8 GB2.2 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q4_K_S
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q4_K_S
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q4_K_S.gguf" --local-dir Qwen3-4B-Base-GGUF
Q4_K_M2.3 GBfile size4.0 GB2.3 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q4_K_M
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q4_K_M
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q4_K_M.gguf" --local-dir Qwen3-4B-Base-GGUF
Q5_K_S2.6 GBfile size4.3 GB2.6 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q5_K_S
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q5_K_S
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q5_K_S.gguf" --local-dir Qwen3-4B-Base-GGUF
Q5_K_M2.7 GBfile size4.3 GB2.7 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q5_K_M
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q5_K_M
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q5_K_M.gguf" --local-dir Qwen3-4B-Base-GGUF
Q6_K3.1 GBfile size4.7 GB3.1 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q6_K
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q6_K
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q6_K.gguf" --local-dir Qwen3-4B-Base-GGUF
Q8_04.0 GBfile size5.6 GB4.0 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:Q8_0
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:Q8_0
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.Q8_0.gguf" --local-dir Qwen3-4B-Base-GGUF
F167.5 GBfile size9.1 GB7.5 + 1.1 + 0.50
Commands
llama.cpp
llama-server -hf mradermacher/Qwen3-4B-Base-GGUF:F16
Ollama
ollama run hf.co/mradermacher/Qwen3-4B-Base-GGUF:F16
Download
hf download mradermacher/Qwen3-4B-Base-GGUF --include "Qwen3-4B-Base.f16.gguf" --local-dir Qwen3-4B-Base-GGUF

MLX for Apple silicon

FormatSizeMemory neededRepo
MLX 4bitjohnlakness/Qwen3-4B-Base-mlx-4Bit2.1 GBfile size3.7 GB2.1 + 1.1 + 0.50Download page

Original weights size from the safetensors header, for vLLM, SGLang, Transformers

FormatSizeMemory neededRepo
BF16original weights7.5 GBfile size9.1 GB7.5 + 1.1 + 0.50
hf download Qwen/Qwen3-4B-Base --local-dir Qwen3-4B-Base
Other conversions (8; sizes not read)

The Hub tags 65 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.

Run it by API

Hugging Face Inference Providers, fetched 19 Nov 2026, 13:58 UTC

No provider on Hugging Face Inference Providers serves this model (as of 19 Nov 2026, 13:58 UTC). Other hosts may; we only list what the router reports.

Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.

Fine-tunes of this model

from our snapshot of the Hub

Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.

Planned for this page Not built yet

Try it in private chat
A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
Fine-tune it
Fine-tuning a few base models from this page and publishing the result here.
Creators and earnings
Pages for people who publish fine-tunes and adapters of this model, and what they earn.
Fund and bounties
Bounties for new open models, paid from a public fund. How the fund will work.
Contribute compute
How to give compute to open model work, with Akash.

Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.