models · google

Gemma 3 4B IT

Google Gemma 3 compact 4B instruct model. Multimodal (text + image). Designed for edge and low-resource deployments.

What this registry has checked

Quarantine. We hold the publisher’s own published hash for these weights and have recorded it against this entry. We have not downloaded and hashed the weights ourselves, so the hash is the publisher’s claim, not our measurement.

Weight file SHA-256	125a8f8f23ceea7ab698640c72cd6de9fe05cb4e0983d185987c2c3c628ed1b3
Method	hf-published-sha256

Re-check this yourself → Source → Full record →

What is recorded

Publisher	google
Origin	United States
Licence	Gemma
Parameters	4,300,079,472 (4.3B)
Context window	128K
Modality	text+image->text
Native precision	BF16
Pulls recorded here	720,019
Downloads reported upstream	1,709,353

Measured on real hardware

None of these runs are ours. Each figure is the reporter’s, linked to their own evidence, under their own trust label. The number shown is single-stream decode — one user, concurrency 1 — because aggregate throughput across many users is a different quantity and reads as far faster than anyone will see.

Reported to local-ai-registry

Hardware	Engine	Build measured	Decode, 1 stream	Label	Reported by
Apple M3 Ultra 512GB	llama.cpp	Q4_K_M	127 tok/s @ 8,192	candidate	localmaxxing
GeForce RTX 3060	llama.cpp	Q4_K_M	97.0 tok/s @ 4,608	candidate	localmaxxing

From local-ai-registry (MIT), commit 124959d: 2 runs across 2 machines, 0 validated — model revision and runtime pinned, launch accepted — and 2 candidate, their label for useful evidence without a reproducible-launch promise.

Elsewhere in the registry

All models → Browse with filters → What will run on my hardware? →

Registry record as of 2026-07-25. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.