Model Follower.

Up to date

Rebuilt
2026-10-25 06:54 UTC
Release radar
117 on watch
Hub snapshot
2026-10-25
Re-checked
daily · every number sourced and dated

Models › gemma-4 family

gemma-4-E4B-it-assistant

Hugging Face ↗ Follow changes

google Apache 2.0 out 6 months ago

78.8M parameters, several kinds of input and output.

Verified 2026-10-25 Rank #760 in LLMs · score 71.3 method v1 Sources 3 Downloads 30d 11.6k · #502 of 619 tracked

Can I use it?Apache 2.0Permissive; patent grantlicence file
Do people use it?11.6kdownloads, 30 days#502 of 619 we track · below the median135 likesHugging Face

Worth reading

Picked by rule from 27 tracked pieces: the lab's own post, the most shared, a field report, the latest.

Most discussed · 868 HN pts

Gemma 4 on iPhone

apps.apple.com 2026-04-05 868 HN pts · 234 comments about the gemma 4 family family

25 best of 27 tracked

The 2 picks above, plus these 23. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.

  1. CodeShow HN: Open-source engine running Gemma 4 26B in 2 GB RAM on any M-series Macgithub.com · 2026-07-29
  2. ArticleGemma 4 12B: A unified, encoder-free multimodal modelblog.google · 2026-06-03 · official
  3. ArticleRunning Gemma 4 26B at 5 tokens/sec on a 13-year-old Xeon with no GPUneomindlabs.com · 2026-07-15
  4. CodeShow HN: Cactus Hybrid: We taught Gemma 4 to know when it's wronggithub.com · 2026-07-22 · 191 HN pts
  5. ArticleGemma 4 QAT models: Optimizing compression for mobile and laptop efficiencyblog.google · 2026-06-05 · 406 HN pts · official
  6. ArticleAccelerating Gemma 4: faster inference with multi-token prediction draftersblog.google · 2026-05-05 · 687 HN pts · official
  7. ArticleGoogle releases Gemma 4 open modelsdeepmind.google · 2026-04-02 · 1,812 HN pts · official
  8. ArticleFable 5 pushed Gemma 4 to 255 tok/s on WebGPUxcancel.com · 2026-06-18 · 48 HN pts
  9. ArticleGoogle Gemma 4 Runs Natively on iPhone with Full Offline AI Inferencegizmoweek.com · 2026-04-15 · 303 HN pts
  10. CodeApril 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac minigist.github.com · 2026-04-03
  11. ArticleShow HN: Determinstic LLM inference for lowest price Gemma 4, with Windows XPtokendelivery.ai · 2026-09-12 · 9 HN pts
  12. ArticleI ran Gemma 4 as a local model in Codex CLIblog.danielvaughan.com · 2026-04-12 · 285 HN pts
  13. ArticleRunning Gemma 4 locally with LM Studio's new headless CLI and Claude Codeai.georgeliu.com · 2026-04-05
  14. ArticleGemma 4 on Cerebras - The Fastest Inference Is Now Multimodalcerebras.ai · 2026-06-30
  15. CodeShow HN: Gemma 4 Multimodal Fine-Tuner for Apple Silicongithub.com · 2026-04-07
  16. ArticleShow HN: Prompt-to-Excalidraw demo with Gemma 4 E2B in the browser (3.1GB)teamchong.github.io · 2026-04-19
  17. CodeShow HN: Local privacy-first Microsoft Recall alternative with Gemma 4github.com · 2026-07-04
  18. ArticleGemma 4: Byte for byte, the most capable open modelsblog.google · 2026-04-02 · official
  19. ArticleA Visual Guide to Gemma 4 12Bnewsletter.maartengrootendorst.com · 2026-06-03
  20. CodeShow HN: Local LLM code-generation with Gemma 4 e2B via JSON AST to Clojuregithub.com · 2026-05-19
  21. ArticleShow HN: Running Gemma-4 26B at 124 tokens/SEC on a CPU, no GPUapeg.dev · 2026-06-30
  22. ArticleShow HN: I made a Gemma 4 Mac app that names screenshots with local AIsnapname.app · 2026-05-31
  23. ArticleGoogle's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAMarstechnica.com · 2026-06-03

Where it's used

Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.

No API provider serves it, so its downloads are people running it themselves: 9 quantized builds.

  • As a service0no API provider listed
  • On own hardware9quantized builds
  • Built on55 fine-tunes · 0 adapters · 0 merges
  • In apps0Hugging Face Spaces

Will it fit?

Weights only, an estimate: leave room for the KV cache. Method

Fits a 16 GB card at FP8

About 0.1 GB at FP8, 0.2 GB at BF16: weights only, so leave room for the KV cache.

16 GB card64 GB Mac2×80 GB8×80 GBFP8: about 0.1 GB of weights (estimate; KV cache and activations extra)FP80.1 GBBF16: about 0.2 GB of weights (estimate; KV cache and activations extra)BF160.2 GB
16 GB card64 GB Mac2×80 GB8×80 GBFP8: about 0.1 GB of weights (estimate; KV cache and activations extra)FP80.1 GBBF16: about 0.2 GB of weights (estimate; KV cache and activations extra)BF160.2 GB

Estimate params × bytes per param × 1.2 · how we estimate

The story so far

Weights, every dated article, and repo commits on one rail.

Most coverage came in the first two weeks

Teal is the weights (2026-04-23), amber triangles are 10 dated articles (each links to it).

MayJunJulAugSepOctShow HN: Cactus Hybrid: We taught Gemma 4 to know when it's wrong · 2026-07-22Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiency · 2026-06-05Accelerating Gemma 4: faster inference with multi-token prediction drafters · 2026-05-05Google releases Gemma 4 open models · 2026-04-02Gemma 4 E2B inference in 700 lines of C · 2026-08-27Fable 5 pushed Gemma 4 to 255 tok/s on WebGPU · 2026-06-18Gemma 4 on iPhone · 2026-04-05Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference · 2026-04-15Show HN: Determinstic LLM inference for lowest price Gemma 4, with Windows XP · 2026-09-12I ran Gemma 4 as a local model in Codex CLI · 2026-04-12Weights 2026-04-23 · 2026-04-23Weights 2026-04-23now
MayJunJulAugSepOctShow HN: Cactus Hybrid: We taught Gemma 4 to know when it's wrong · 2026-07-22Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiency · 2026-06-05Accelerating Gemma 4: faster inference with multi-token prediction drafters · 2026-05-05Google releases Gemma 4 open models · 2026-04-02Gemma 4 E2B inference in 700 lines of C · 2026-08-27Fable 5 pushed Gemma 4 to 255 tok/s on WebGPU · 2026-06-18Gemma 4 on iPhone · 2026-04-05Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference · 2026-04-15Show HN: Determinstic LLM inference for lowest price Gemma 4, with Windows XP · 2026-09-12I ran Gemma 4 as a local model in Codex CLI · 2026-04-12Weights 2026-04-23 · 2026-04-23Weights 2026-04-23now

What people built on it

14 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.

14 derivatives on the Hub, most of them quantized

From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.

9 quantized (Hugging Face model tree)5 fine-tunes (Hugging Face model tree)9 quantized5 fine-tunes
9 quantized (Hugging Face model tree)5 fine-tunes (Hugging Face model tree)9 quantized5 fine-tunes
GGUF AtomicChat/gemma-4-E4B-it-assistant-GGUF for llama.cpp, Ollama, LM Studio 35.5k downloads MLX mlx-community/gemma-4-E4B-it-assistant-bf16 for Apple silicon 14.9k downloads

All variants →

Reference

Everything else, folded: the full tables, every source.

Use-case evidence

3 use cases

What it is good at 3 use cases

Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.

Facts

15 fields

Listed in LLMs

Facts

Pipeline tagany-to-any
Librarytransformers
ArchitecturesGemma4AssistantForCausalLM
Model typegemma4_assistant
Params by dtypeBF16: 78,517,764 · I64: 262,144
Size bandS
MoEno
PapersarXiv:2607.02770
LicenseApache 2.0 (apache-2.0)
Hub snapshot2026-10-25

Hardware fit

3 estimates

Hardware fit ESTIMATE

Weights-only memory for gemma-4-E4B-it-assistant: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.

BF16 (2 bytes/param)0.2 GB · Edge (<4 GB)
FP8 (1 byte/param)0.1 GB · Edge (<4 GB)
INT4 (0.5 bytes/param)0.0 GB · Edge (<4 GB)

Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.

All variants

14 variants

Lineage

Original model Family: google/gemma-4 14 derivatives on the Hub

quantized: 9finetune: 5

Top variants (14 shown, ranked by downloads)

Relation
Quant
14 rows
Variant Relation Quant Params Band License Author Likes Downloads
AtomicChat/gemma-4-E4B-it-assistant-GGUF quantized GGUF — — apache-2.0 AtomicChat 26 35.5k
mlx-community/gemma-4-E4B-it-assistant-bf16 finetune -bf16 78.8M S gemma mlx-community 12 14.9k
google/gemma-4-E4B-it-qat-q4_0-unquantized-assistant finetune q4_0 78.8M S apache-2.0 google 21 11.1k
cafkafk/gemma-4-E4B-it-assistant-GGUF-noimatrix quantized GGUF — — gemma cafkafk 3 5.1k
matrixportalx/gemma-4-E4B-it-assistant-MTP-GGUF quantized GGUF — — gemma matrixportalx 1 2.2k
Qwe1325/gemma-4-E4B-it-assistant-gguf quantized gguf — — — Qwe1325 0 1.6k
dahara1/gemma-4-E4B-it-assistant_gguf quantized gguf — — apache-2.0 dahara1 0 767
mags0ft/gemma-4-E4B-it-assistant-gguf quantized gguf — — apache-2.0 mags0ft 3 610
srmiles/gemma-4-E4B-it-assistant-GGUF quantized GGUF — — apache-2.0 srmiles 1 522
NeoRoth/gemma-4-e4b-it-assistant-mlx finetune mlx 78.8M S gemma NeoRoth 0 169
LeadingPoint/gemma-4-E4B-it-assistant-w4a16 quantized -w4a16 78.8M S apache-2.0 LeadingPoint 0 142
yassaalfy/gemma-4-E4B-it-qat-q4_0-unquantized-assistant finetune q4_0 78.8M S apache-2.0 yassaalfy 0 58
OsGo/gemma-4-E4B-it-qat-q4_0-unquantized-assistant finetune q4_0 78.8M S apache-2.0 OsGo 0 35
RASMUS/gemma-4-E4B-it-mtp-ONNX quantized — — — gemma RASMUS 0 0

Repository changelog

1 change
changed in the last 7 daysNew on the watch list: 2026-10-23all changes

Articles and coverage

27 pieces

Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.

Official3

Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiency
blog.google2026-06-05officialauthorityHN front pageabout the gemma 4 familyHN 406 pts · 130 comments714 words
Google released new Gemma 4 checkpoints optimized with Quantization-Aware Training to improve efficiency on edge devices and consumer GPUs. The E2B model achieves a 1GB memory footprint using a mobile-specialized schema that applies targeted 2-bit quantization to token generation layers. Developers can access GGUF formats ready for llama.cpp and MTP QAT checkpoints to preserve inference speedups while running locally.written, reviewed, cited
  • The mobile-specialized schema pre-calculates activation scaling settings during training to reduce workload on mobile chips.
  • Targeted 2-bit quantization compresses token generation parts while keeping core reasoning layers at higher precision.
  • Deploying only needed modalities allows the Gemma 4 E2B text-only model to require less than 1 GB of memory.
Accelerating Gemma 4: faster inference with multi-token prediction drafters
blog.google2026-05-05officialauthorityHN front pageabout the gemma 4 familyHN 687 pts · 330 comments796 words
Google released Multi-Token Prediction (MTP) drafters for the Gemma 4 family under the same open-source Apache 2.0 license as Gemma 4. These drafters deliver up to a 3x speedup for models like the 26B MoE and 31B Dense versions without quality degradation. Benchmarks on hardware such as the NVIDIA RTX PRO 6000 show that increasing batch sizes to 4 to 8 unlocks up to a ~2.2x speedup locally.written, reviewed, cited
  • The MTP drafter shares the target model's KV cache to avoid recalculating context the larger model has already processed.
  • For E2B and E4B edge models, an efficient clustering technique in the embedder accelerates generation where logit calculation is a bottleneck.
  • The 26B mixture-of-experts model presents unique routing challenges at a batch size of 1 on Apple Silicon.
Google releases Gemma 4 open models
deepmind.google2026-04-02officialauthorityHN front pageabout the gemma 4 familyHN 1,812 pts · 474 comments253 words

Analysis7

Show HN: Cactus Hybrid: We taught Gemma 4 to know when it's wrong
github.com2026-07-22authorityHN front pageabout the gemma 4 familyHN 191 pts · 44 comments714 words
Cactus Hybrid introduces on-device models that score every answer with a confidence between 0 and 1 to decide when to hand off to a bigger model. The Gemma 4 E2B Hybrid model matches Gemini 3.1 Flash-Lite performance by routing only 15–55% of queries to the cloud while running the rest locally. It is MIT-licensed and achieves a 0.770 AUROC on MMLU, with a probe trained on zero audio data yet reaching 0.79–0.88 AUROC on four audio benchmarks.written, reviewed, cited
  • Weights on the meta device crash the confidence read because probes score generations outside the module forward path.
  • The probe is a patch you compile into the llama.cpp engine rather than a standalone library component.
  • Transformers version 5.14 and above cause a segfault on this specific checkpoint.
Gemma 4 E2B inference in 700 lines of C
github.com2026-08-27authorityabout the gemma 4 familyHN 18 pts · 0 comments722 words
Fable 5 pushed Gemma 4 to 255 tok/s on WebGPU
xcancel.com2026-06-18about the gemma 4 familyHN 48 pts · 22 comments
Gemma 4 on iPhone
apps.apple.com2026-04-05HN front pageabout the gemma 4 familyHN 868 pts · 234 comments
Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference
gizmoweek.com2026-04-15HN front pageabout the gemma 4 familyHN 303 pts · 187 comments364 words
Show HN: Determinstic LLM inference for lowest price Gemma 4, with Windows XP
tokendelivery.ai2026-09-12about the gemma 4 familyHN 9 pts · 1 comments
I ran Gemma 4 as a local model in Codex CLI
blog.danielvaughan.com2026-04-12HN front pageabout the gemma 4 familyHN 285 pts · 116 comments
All coverage tracked for this model (27) — append-only, nothing is ever removed
TitleSourcePublishedFirst seenSignals
Show HN: Determinstic LLM inference for lowest price Gemma 4, with Windows XPtokendelivery.ai2026-09-122026-10-09hn
Gemma 4 E2B inference in 700 lines of Cgithub.com2026-08-272026-10-09hn
Show HN: Open-source engine running Gemma 4 26B in 2 GB RAM on any M-series Macgithub.com2026-07-292026-10-09hn
Show HN: Cactus Hybrid: We taught Gemma 4 to know when it's wronggithub.com2026-07-222026-10-09hn
Running Gemma 4 26B at 5 tokens/sec on a 13-year-old Xeon with no GPUneomindlabs.com2026-07-152026-10-09hn
Show HN: Local privacy-first Microsoft Recall alternative with Gemma 4github.com2026-07-042026-10-09hn
Show HN: Running Gemma-4 26B at 124 tokens/SEC on a CPU, no GPUapeg.dev2026-06-302026-10-09hn
Gemma 4 on Cerebras - The Fastest Inference Is Now Multimodalcerebras.ai2026-06-302026-10-09hn
Fable 5 pushed Gemma 4 to 255 tok/s on WebGPUxcancel.com2026-06-182026-10-09hn
Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiencyblog.google official2026-06-052026-10-09hn
Google's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAMarstechnica.com2026-06-032026-10-09hn
A Visual Guide to Gemma 4 12Bnewsletter.maartengrootendorst.com2026-06-032026-10-09hn
Gemma 4 12B: A unified, encoder-free multimodal modelblog.google official2026-06-032026-10-09hn
Show HN: I made a Gemma 4 Mac app that names screenshots with local AIsnapname.app2026-05-312026-10-09hn
Show HN: Local LLM code-generation with Gemma 4 e2B via JSON AST to Clojuregithub.com2026-05-192026-10-09hn
Accelerating Gemma 4: faster inference with multi-token prediction draftersblog.google official2026-05-052026-10-09hn
Show HN: Prompt-to-Excalidraw demo with Gemma 4 E2B in the browser (3.1GB)teamchong.github.io2026-04-192026-10-09hn
Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inferencegizmoweek.com2026-04-152026-10-09hn
Show HN: I benchmarked Gemma 4 E2B – the 2B model beat the 12B on multi-turnaiexplr.com2026-04-132026-10-09hn
I ran Gemma 4 as a local model in Codex CLIblog.danielvaughan.com2026-04-122026-10-09hn
A Visual Guide to Gemma 4newsletter.maartengrootendorst.com2026-04-092026-10-09hn
Show HN: Gemma 4 Multimodal Fine-Tuner for Apple Silicongithub.com2026-04-072026-10-09hn
Gemma 4 on iPhoneapps.apple.com2026-04-052026-10-09hn
Running Gemma 4 locally with LM Studio's new headless CLI and Claude Codeai.georgeliu.com2026-04-052026-10-09hn
April 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac minigist.github.com2026-04-032026-10-09hn
Google releases Gemma 4 open modelsdeepmind.google official2026-04-022026-10-09hn
Gemma 4: Byte for byte, the most capable open modelsblog.google official2026-04-022026-10-09hn

Papers

1 paper

Papers 1

arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).

Similar models

3 models

Similar models

Sources

3 sources

Sources

Hub snapshot: 2026-10-25 · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-09 02:04 UTC · Last verified 2026-10-25

Follow gemma-4-E4B-it-assistant

Watch this model

Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.