Skip to content
Weights Not Walls
Menu

google

medgemma-4b-it

Reviewedgated on the Hubaccepts images
Size
4.3B parameterscounted from the safetensors header on the Hub
Licence
custom, read it, Health AI Developer Foundations Terms of UseOur reading, check the licence: A custom licence we have not read for this model: other makers use the same name for different texts. Read the model's own licence file before you use it.Licence text
On the Hub since
19 May 2025the date the repo was created
Context length
Not known
Fine-tune
of google/medgemma-4b-ptGemma3ForConditionalGeneration
On Hugging Face
google/medgemma-4b-it469k downloads in the last 30 days, 1.2k 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: 0.27 GB, worked out from the model's config (jesusoctavioas/medgemma-4b-it-original-mlx-4Bit), 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

unsloth/medgemma-4b-it-GGUF (converted by unsloth), 29k downloads. Sizes are the real files on the Hub, all parts added up. For image input you also need the projector file mmproj-BF16.gguf (0.79 GB), not counted below.

FormatSizeMemory neededGet it
UD-IQ3_XXS1.6 GBfile size2.4 GB1.6 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:UD-IQ3_XXS
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:UD-IQ3_XXS
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-UD-IQ3_XXS.gguf" --local-dir medgemma-4b-it-GGUF
Q3_K_S1.8 GBfile size2.6 GB1.8 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q3_K_S
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q3_K_S
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q3_K_S.gguf" --local-dir medgemma-4b-it-GGUF
Q3_K_M2.0 GBfile size2.7 GB2.0 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q3_K_M
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q3_K_M
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q3_K_M.gguf" --local-dir medgemma-4b-it-GGUF
UD-Q3_K_XL2.0 GBfile size2.8 GB2.0 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:UD-Q3_K_XL
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:UD-Q3_K_XL
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-UD-Q3_K_XL.gguf" --local-dir medgemma-4b-it-GGUF
IQ4_XS2.1 GBfile size2.9 GB2.1 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:IQ4_XS
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:IQ4_XS
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-IQ4_XS.gguf" --local-dir medgemma-4b-it-GGUF
IQ4_NL2.2 GBfile size3.0 GB2.2 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:IQ4_NL
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:IQ4_NL
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-IQ4_NL.gguf" --local-dir medgemma-4b-it-GGUF
Q4_02.2 GBfile size3.0 GB2.2 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q4_0
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q4_0
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q4_0.gguf" --local-dir medgemma-4b-it-GGUF
Q4_K_S2.2 GBfile size3.0 GB2.2 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q4_K_S
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q4_K_S
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q4_K_S.gguf" --local-dir medgemma-4b-it-GGUF
Q4_K_M2.3 GBfile size3.1 GB2.3 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q4_K_M
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q4_K_M
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q4_K_M.gguf" --local-dir medgemma-4b-it-GGUF
UD-Q4_K_XL2.4 GBfile size3.1 GB2.4 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:UD-Q4_K_XL
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:UD-Q4_K_XL
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-UD-Q4_K_XL.gguf" --local-dir medgemma-4b-it-GGUF
Q4_12.4 GBfile size3.2 GB2.4 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q4_1
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q4_1
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q4_1.gguf" --local-dir medgemma-4b-it-GGUF
Q5_K_S2.6 GBfile size3.3 GB2.6 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q5_K_S
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q5_K_S
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q5_K_S.gguf" --local-dir medgemma-4b-it-GGUF
Q5_K_M2.6 GBfile size3.4 GB2.6 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q5_K_M
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q5_K_M
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q5_K_M.gguf" --local-dir medgemma-4b-it-GGUF
UD-Q5_K_XL2.6 GBfile size3.4 GB2.6 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:UD-Q5_K_XL
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:UD-Q5_K_XL
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-UD-Q5_K_XL.gguf" --local-dir medgemma-4b-it-GGUF
Q6_K3.0 GBfile size3.7 GB3.0 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q6_K
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q6_K
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q6_K.gguf" --local-dir medgemma-4b-it-GGUF
UD-Q6_K_XL3.3 GBfile size4.1 GB3.3 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:UD-Q6_K_XL
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:UD-Q6_K_XL
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-UD-Q6_K_XL.gguf" --local-dir medgemma-4b-it-GGUF
Q8_03.8 GBfile size4.6 GB3.8 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:Q8_0
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:Q8_0
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-Q8_0.gguf" --local-dir medgemma-4b-it-GGUF
UD-Q8_K_XL4.8 GBfile size5.6 GB4.8 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:UD-Q8_K_XL
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:UD-Q8_K_XL
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-UD-Q8_K_XL.gguf" --local-dir medgemma-4b-it-GGUF
BF167.2 GBfile size8.0 GB7.2 + 0.27 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/medgemma-4b-it-GGUF:BF16
Ollama
ollama run hf.co/unsloth/medgemma-4b-it-GGUF:BF16
Download
hf download unsloth/medgemma-4b-it-GGUF --include "medgemma-4b-it-BF16.gguf" --local-dir medgemma-4b-it-GGUF

MLX for Apple silicon

FormatSizeMemory neededRepo
MLX 4bitjesusoctavioas/medgemma-4b-it-original-mlx-4Bit2.0 GBfile size2.8 GB2.0 + 0.27 + 0.50Download page

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

FormatSizeMemory neededRepo
BF16original weights8.0 GBfile size8.8 GB8.0 + 0.27 + 0.50
hf download google/medgemma-4b-it --local-dir medgemma-4b-it
Other conversions (10; sizes not read)

The Hub tags 57 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.

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.