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-3 family

gemma-3-1b-pt

Hugging Face ↗ Follow changes

google Gemma Terms of Use out 20 months ago

1000M parameters, text in, text out, 32K tokens of context. 1 provider serves it.

Verified 2026-10-25 Rank #428 in LLMs · score 78.6 method v1 Sources 3 Downloads 30d 36.8k · #406 of 619 tracked

Changed this week: 2 new quantized builds on the Hub (24 in all).2026-10-24

Can I run it?0.6GB at INT4Estimatefits a 16 GB cardmethod
Can I use it?Gemma Terms of UseGoogle's terms + prohibited-use policy; must pass terms downstreamlicence file2026-10-24
Do people use it?36.8kdownloads, 30 days#406 of 619 we track · below the median+0 likes, downloads −4% this week218 likesHugging Face2026-10-24

Worth reading

Picked by rule from 26 tracked pieces: the lab's own post, the most shared, a field report, the latest.

Most discussed · 488 HN pts

Gemma 3 Technical Report [pdf]

storage.googleapis.com 2025-03-12 488 HN pts · 245 comments about the gemma 3 family family
Latest · 4 months ago

Show HN: Gemma 3 inference in pure C++ with Metal acceleration

MetalChat is a Metal-accelerated C++ framework and command line interpreter for inference of Meta Llama, Google Gemma, and Liquid models.

github.com 2026-07-04 6 HN pts · 2 comments 193 words about the gemma 3 family family

25 best of 26 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. CodeOpen-jev: One-pass option scoring with Gemma 3 4B, similar to jevgithub.com · 2026-09-17
  2. ArticleGemma 3 270M: Compact model for hyper-efficient AIdevelopers.googleblog.com · 2025-08-14 · official
  3. ArticleGemma 3 QAT Models: Bringing AI to Consumer GPUsdevelopers.googleblog.com · 2025-04-18 · 602 HN pts · official
  4. ArticleIntroducing Gemma 3ndevelopers.googleblog.com · 2025-06-26 · 405 HN pts · official
  5. ArticleGemma 3n preview: Mobile-first AIdevelopers.googleblog.com · 2025-05-20 · 441 HN pts · official
  6. ArticleGemma3 Function Callingai.google.dev · 2025-03-23 · 202 HN pts · official
  7. ArticleFine-tune Google's Gemma 3unsloth.ai · 2025-03-19 · 226 HN pts
  8. CodeGemma 3 270M re-implemented in pure PyTorch for local tinkeringgithub.com · 2025-08-20
  9. CodeI have written gemma3 inference in pure Cgithub.com · 2026-01-26 · 65 HN pts
  10. ArticleGemma3 – The current strongest model that fits on a single GPUollama.com · 2025-03-12 · 252 HN pts
  11. ArticleGoogle calls Gemma 3 the most powerful AI model you can run on one GPUtheverge.com · 2025-03-20 · 127 HN pts
  12. ArticleGemma 3: Google's new multimodal modelsai.google.dev · 2025-03-12 · official
  13. ArticleGemma 3: The most capable model you can run on a single GPU or TPUblog.google · 2025-03-12 · official
  14. ArticleOwn your AI: Learn how to fine-tune Gemma 3 270M and run it on-devicedevelopers.googleblog.com · 2025-10-10 · official
  15. ArticleGemma 3 QAT Modelssimonwillison.net · 2025-04-19
  16. DiscussionGemma 3n Architectural Innovations – Speculation and poking around in the modelold.reddit.com · 2025-05-25
  17. ArticleGemma 3n available in the open-source ecosystemhuggingface.co · 2025-06-26
  18. ArticleGemma 3n: The Developer Guidesimonwillison.net · 2025-06-27
  19. ArticleLandrecords – cheap nationwide parcel dataset standardized using gemma3landrecords.us · 2025-08-20
  20. CodeA C++ library to efficiently run Gemma-3N across various platformgithub.com · 2025-06-11
  21. ArticleSelectively reducing eval awareness and murder in Gemma 3 27B via steeringlesswrong.com · 2026-03-12
  22. ArticleI Gave Up Trying to Run Gemma 3 27B on AWS P3dn.24xlargemedium.com · 2025-08-11
  23. ArticleGoogle's Gemma 3 is an open source, multimodal single-GPU AI 128K context windowarstechnica.com · 2025-03-13

Where to run it

Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.

#Cheapest live rowIn /1MOut /1MContextQuantSeen
1 Featherless — — — — Oct 25

All 1 provider →

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.

Served by 1 API provider; 24 quantized builds for local use; a big base for builders (499 fine-tunes, adapters and merges).

  • As a service1Featherless
  • On own hardware24quantized builds
  • Built on499427 fine-tunes · 62 adapters · 10 merges
  • In apps20Hugging Face Spaces

Top Spaces running it: stats-powered-ai/StatDetectLLM ♥ 9 · Jasmineame/detect-LLM-tokens ♥ 4 · oopere/optipfair-bias-analyzer ♥ 3 · GiacomoDeLuca/orrery-demo ♥ 3

Will it fit?

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

Fits a 16 GB card at INT4

About 0.6 GB at INT4, 1.2 GB at FP8, 2.4 GB at BF16: weights only, so leave room for the KV cache.

16 GB card64 GB Mac2×80 GB8×80 GBINT4: about 0.6 GB of weights (estimate; KV cache and activations extra)INT40.6 GBFP8: about 1.2 GB of weights (estimate; KV cache and activations extra)FP81.2 GBBF16: about 2.4 GB of weights (estimate; KV cache and activations extra)BF162.4 GB
16 GB card64 GB Mac2×80 GB8×80 GBINT4: about 0.6 GB of weights (estimate; KV cache and activations extra)INT40.6 GBFP8: about 1.2 GB of weights (estimate; KV cache and activations extra)FP81.2 GBBF16: about 2.4 GB of weights (estimate; KV cache and activations extra)BF162.4 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.

10 dated articles and 6 repo commits so far

Teal is the weights (2025-02-20), amber triangles are 10 dated articles (each links to it), grey ticks are repo commits.

MarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2025-03-21: 1 repo commit2025-03-19: 1 repo commit2025-03-14: 2 repo commits2025-03-12: 2 repo commitsShow HN: Gemma 3 inference in pure C++ with Metal acceleration · 2026-07-04Gemma 3 QAT Models: Bringing AI to Consumer GPUs · 2025-04-18Introducing Gemma 3n · 2025-06-26Gemma 3n preview: Mobile-first AI · 2025-05-20Gemma3 Function Calling · 2025-03-23Fine-tune Google's Gemma 3 · 2025-03-19Gemma 3 Technical Report [pdf] · 2025-03-12I have written gemma3 inference in pure C · 2026-01-26Gemma3 – The current strongest model that fits on a single GPU · 2025-03-12Google calls Gemma 3 the most powerful AI model you can run on one GPU · 2025-03-20Weights 2025-02-20 · 2025-02-20Weights 2025-02-20now · 1 provider
MarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2025-03-21: 1 repo commit2025-03-19: 1 repo commit2025-03-14: 2 repo commits2025-03-12: 2 repo commitsShow HN: Gemma 3 inference in pure C++ with Metal acceleration · 2026-07-04Gemma 3 QAT Models: Bringing AI to Consumer GPUs · 2025-04-18Introducing Gemma 3n · 2025-06-26Gemma 3n preview: Mobile-first AI · 2025-05-20Gemma3 Function Calling · 2025-03-23Fine-tune Google's Gemma 3 · 2025-03-19Gemma 3 Technical Report [pdf] · 2025-03-12I have written gemma3 inference in pure C · 2026-01-26Gemma3 – The current strongest model that fits on a single GPU · 2025-03-12Google calls Gemma 3 the most powerful AI model you can run on one GPU · 2025-03-20Weights 2025-02-20 · 2025-02-20Weights 2025-02-20now · 1 provider

What people built on it

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

523 derivatives on the Hub, most of them fine-tunes

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

24 quantized (Hugging Face model tree)427 fine-tunes (Hugging Face model tree)10 merges (Hugging Face model tree)62 adapters (Hugging Face model tree)24 quantized427 fine-tunes10 merges62 adapters
24 quantized (Hugging Face model tree)427 fine-tunes (Hugging Face model tree)10 merges (Hugging Face model tree)62 adapters (Hugging Face model tree)24 quantized427 fine-tunes10 merges62 adapters
GGUF mradermacher/gemma-3-1b-pt-i1-GGUF for llama.cpp, Ollama, LM Studio 14.8k downloads MLX mlx-community/gemma-3-1b-pt-4bit for Apple silicon 10.3k downloads

All variants →

Reference

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

Benchmarks

17 rows

Benchmarks 17 rows

measured third-party leaderboard rowauthor-reported number from the model card or eval files (17)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.
BenchmarkValueSetting / notesReported byDate
author-reported ARC Challenge · Chat and follow instructions38.4Gemma 3 PT 1B, 25-shotmodel author2026-10-07
author-reported BBH · Reason through hard problems28.4Gemma 3 PT 1B, few-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions63.2Gemma 3 PT 1B, 0-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions73.8Gemma 3 PT 1B, 0-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions48.9Gemma 3 PT 1B, 0-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions39.8Gemma 3 PT 1B, 5-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions9.48Gemma 3 PT 1B, 5-shotmodel author2026-10-07
author-reported Commonsense QA sets · Chat and follow instructions73Gemma 3 PT 1B, 0-shotmodel author2026-10-07
author-reported DROP42.4Gemma 3 PT 1B, 1-shotmodel author2026-10-07
author-reported ECLeKTic4.69Gemma 3 PT 1Bmodel author2026-10-07
author-reported FLORES · Work across languages29.5Gemma 3 PT 1Bmodel author2026-10-07
author-reported Global-MMLU · Work across languages24.9Gemma 3 PT 1Bmodel author2026-10-07
author-reported HellaSwag · Chat and follow instructions62.3Gemma 3 PT 1B, 10-shotmodel author2026-10-07
author-reported IndicGenBench41.4Gemma 3 PT 1Bmodel author2026-10-07
author-reported MGSM · Work across languages2.04Gemma 3 PT 1Bmodel author2026-10-07
author-reported Winogrande · Chat and follow instructions58.2Gemma 3 PT 1B, 5-shotmodel author2026-10-07
author-reported XQuAD (all)43.9Gemma 3 PT 1Bmodel author2026-10-07

What the card says

2 cited sentences

Summary generated, cited

Gemma 3 1B PT is a 1000M parameter multimodal model designed for text generation and image understanding.[1][2] This 1B size variant supports a 32K token input context.[3]

  1. parameters human
  2. “text generation and image understanding”
  3. “32K tokens for the 1B size”

Generated 2026-10-07 15:09 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.

Use-case evidence

5 use cases

What it is good at 5 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

21 fields

Listed in LLMs

Facts

Pipeline tagtext-generation
Librarytransformers
ArchitecturesGemma3ForCausalLM
Model typegemma3_text
Dtypebfloat16
Params by dtypeBF16: 999,885,952
Size bandS
MoEno
Context length32,768 (config.json max_position_embeddings)
Gatedmanual
PapersarXiv:1905.07830, arXiv:1905.10044, arXiv:1911.11641, arXiv:1904.09728, arXiv:1705.03551, arXiv:1911.01547, arXiv:1907.10641, arXiv:1903.00161, arXiv:2009.03300, arXiv:2304.06364, arXiv:2103.03874, arXiv:2110.14168, arXiv:2311.12022, arXiv:2108.07732, arXiv:2107.03374, arXiv:2210.03057, arXiv:2106.03193, arXiv:1910.11856, arXiv:2502.12404, arXiv:2502.21228, arXiv:2404.16816, arXiv:2104.12756, arXiv:2311.16502, arXiv:2203.10244, arXiv:2404.12390, arXiv:1810.12440, arXiv:1908.02660, arXiv:2312.11805
LicenseGemma Terms of Use (gemma)
Hub snapshot2026-10-25

Servability

vLLM · SGLang

Servability

vLLM: supported · Gemma3ForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Gemma3ForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed

vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.

Hardware fit

3 estimates

Hardware fit ESTIMATE

Weights-only memory for gemma-3-1b-pt: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.

BF16 (2 bytes/param)2.4 GB · Edge (<4 GB)
FP8 (1 byte/param)1.2 GB · Edge (<4 GB)
INT4 (0.5 bytes/param)0.6 GB · Edge (<4 GB)

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

All variants

40 variants

Lineage

Original model Family: google/gemma-3 523 derivatives on the Hub

finetune: 427quantized: 24merge: 10adapter: 62

Top variants (40 shown, ranked by downloads)

Relation
Quant
40 rows
Variant Relation Quant Params Band License Author Likes Downloads
unsloth/gemma-3-1b-pt finetune — 1000M S gemma unsloth 6 95.7k
ai4bharat/Cadence finetune — 1000M S mit ai4bharat 21 83.8k
unsloth/gemma-3-1b-pt-unsloth-bnb-4bit quantized -bnb 1.02B S gemma unsloth 4 36.1k
mradermacher/gemma-3-1b-pt-i1-GGUF quantized GGUF — — gemma mradermacher 1 14.8k
mlx-community/gemma-3-1b-pt-4bit quantized -4bit 1.30B S gemma mlx-community 1 10.3k
unsloth/gemma-3-1b-pt-bnb-4bit quantized -bnb 1000M S gemma unsloth 3 10.2k
alnnahwi/gemma-3-1b-arabic-gec-v1 finetune — 1000M S gemma alnnahwi 9 9.5k
gguf-org/gemma-3-1b-pt-gguf quantized gguf — — gemma gguf-org 3 7.4k
mozilla-ai/gemma-3-1b-it-llamafile finetune — — — gemma mozilla-ai 6 6.4k
jordimas/gemma-3-1b-it finetune — 1000M S gemma jordimas 0 5.8k
mradermacher/gemma-3-1b-pt-GGUF quantized GGUF — — gemma mradermacher 2 4k
scratchyone/anon-1b quantized GGUF 1000M S gemma scratchyone 40 2.1k
DevQuasar-5/google.gemma-3-1b-pt-GGUF quantized GGUF — — — DevQuasar-5 0 4.1k
google/gemma-3-1b-pt-qat-q4_0-gguf quantized q4_0 — — gemma google 17 2.2k
Abdoul-AI/gemma-3-1b-it-GGUF quantized GGUF — — gemma Abdoul-AI 0 1.3k
canbingol/gemma3_1B_base-tr-cpt-only_2nd_stage_data finetune — 1000M S — canbingol 0 1.2k
mlx-community/gemma-3-1b-pt-8bit quantized -8bit 1.30B S gemma mlx-community 0 1.1k
canbingol/gemma3_1B_base-tr-cpt-only_4th_stage_data finetune — 1000M S — canbingol 0 848
mlx-community/gemma-3-1b-pt-bf16 finetune -bf16 1.30B S gemma mlx-community 0 841
jeisonvendetta/blockvault-gemma3-1b finetune — 1000M S — jeisonvendetta 0 771
litert-community/gemma3-1b-ft-text-to-sql finetune — — — gemma litert-community 12 98
GorankLabs/Rank-Embed-1B finetune — 1.00B S apache-2.0 GorankLabs 3 536
MokolIslam/gemma-3-1b-it-qwen3-tool-template finetune — 1000M S gemma MokolIslam 0 679
xiaomi-research/MiLMMT-46-1B-Pretrain finetune — 1.30B S gemma xiaomi-research 4 465
AJ1731715/gemma-3-1b-it-heretic finetune — 1000M S gemma AJ1731715 0 626
harsha2946/gemma-finetuned-sentiment finetune — 268M S — harsha2946 0 563
V4ldeLund/gemma-3-1b-pt-faroese-experiment5 finetune — 1000M S — V4ldeLund 0 556
canbingol/gemma3_1B_base-tr-cpt-only_3rd_stage_data finetune — 1000M S — canbingol 0 460
BidirLM/BidirLM-1B-Base finetune — 1000M S apache-2.0 BidirLM 0 454
tuandunghcmut/gemma-3-1b-it-qwen3-tool-template finetune — 1000M S gemma tuandunghcmut 0 418
rapacious/gemma-3-1b-it finetune — 1000M S gemma rapacious 0 409
Io2007/gemma-3-1b-big merge — 1.11B S — Io2007 0 406
RitwikSehrawat/gemma-3-1b-kbd-q4km adapter GGUF — — apache-2.0 RitwikSehrawat 1 340
Zakeertechie3/gemma-3-1b-pt-Q5_K_M-GGUF quantized GGUF — — gemma Zakeertechie3 0 363
mengsay/Gemma-3-Tonsai-1B-v0.1 finetune — 1000M S gemma mengsay 3 199
coder3101/gemma-3-1b-it-heretic finetune — 1000M S gemma coder3101 0 348
Roman0/gemma-3-1b-it-heretic finetune — 1000M S gemma Roman0 1 293
mlx-community/gemma-3-1b-pt-6bit quantized — 1.30B S gemma mlx-community 0 338
ai-for-good-lab/byol-nya-1b-cpt finetune — 1000M S gemma ai-for-good-lab 0 337
gimmy256/medical-gemma3_1b adapter — — — apache-2.0 gimmy256 0 330

Repository changelog

6 commits
changed in the last 7 daysOther: 2026-10-24all changes

Repository changelog 6 commits

DateCommitSHA
2025-03-21 Update README.md fcf18a2
2025-03-19 Adding top_k and top_p to generation_config.json c87c33c
2025-03-14 Update README.md 1b6b7c0
2025-03-14 Delete model.safetensors.index.json (#6) 5c65e74
2025-03-12 Update README.md fc2bde0
2025-03-12 Super-squash branch 'main' using huggingface_hub 0e7c6f3

All providers

1 provider · 1 endpoint

Where to run 1 providers · 1 endpoints

Hosted endpoints observed for gemma-3-1b-pt, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting.

ProviderModel refContextQuant$/M in$/M outtok/sTTFTToolsStatusObserved
Featherless
hf
 google/gemma-3-1b-pt — — — — — — — live 2026-10-25

Articles and coverage

26 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.

Official4

Gemma 3 QAT Models: Bringing AI to Consumer GPUs
developers.googleblog.com2025-04-18officialauthorityHN front pageabout the gemma 3 familyHN 602 pts · 276 comments1,077 words
Google released Gemma 3 Quantization-Aware Training (QAT) models that incorporate quantization during training to reduce memory usage while maintaining accuracy. Applying QAT on ~5,000 steps reduces the perplexity drop by 54% when quantizing down to Q4_0, allowing the 27B variant to fit in 14.1 GB of VRAM instead of 54 GB. These int4 models run efficiently on consumer hardware like the NVIDIA RTX 3090 and are available via Ollama, llama.cpp, and MLX.written, reviewed, cited
  • Deploying the largest models often requires lower-precision formats like FP8, potentially accepting a performance trade-off for feasibility.
  • The Gemmaverse offers many alternatives using Post-Training Quantization (PTQ) with contributions from members such as Bartowski, Unsloth, and GGML.
Introducing Gemma 3n
developers.googleblog.com2025-06-26officialauthorityHN front pageabout the gemma 3 familyHN 405 pts · 191 comments1,690 words
Gemma 3n is a mobile-first AI model family designed for on-device use with multimodal capabilities. The E2B and E4B versions have raw parameters of 5B and 8B but operate with as little as 2GB and 3GB of memory respectively. It supports 140 languages for text and includes a MobileNet-V5-300M vision encoder that delivers a 13x speedup with quantization, requires 46% fewer parameters, and has a 4x smaller memory footprint.written, reviewed, cited
  • Developers can create custom-sized models between E2B and E4B by adjusting feed forward network hidden dimensions from 8192 to 16384.
  • The Gemma 3n audio encoder processes clips up to 30 seconds at launch but supports arbitrarily long audios with additional training.
Gemma 3n preview: Mobile-first AI
developers.googleblog.com2025-05-20officialauthorityHN front pageabout the gemma 3 familyHN 441 pts · 163 comments904 words
Gemma3 Function Calling
ai.google.dev2025-03-23officialauthorityHN front pageabout the gemma 3 familyHN 202 pts · 32 comments2,144 words

Analysis6

Show HN: Gemma 3 inference in pure C++ with Metal acceleration
github.com2026-07-04authorityabout the gemma 3 familyHN 6 pts · 2 comments193 words
MetalChat is a Metal-accelerated C++ framework and command line interpreter for inference of Meta Llama, Google Gemma, and Liquid models. The library API and CLI are under active development, therefore they may change without any deprecation notice. The framework is distributed under GPLv3 license and can be installed using Homebrew package manager with brew tap ybubnov/metalchat.written, reviewed, cited
Fine-tune Google's Gemma 3
unsloth.ai2025-03-19authorityHN front pageabout the gemma 3 familyHN 226 pts · 78 comments
Gemma 3 Technical Report [pdf]
storage.googleapis.com2025-03-12HN front pageabout the gemma 3 familyHN 488 pts · 245 comments
I have written gemma3 inference in pure C
github.com2026-01-26authorityabout the gemma 3 familyHN 65 pts · 23 comments2,207 words
Gemma3 – The current strongest model that fits on a single GPU
ollama.com2025-03-12HN front pageabout the gemma 3 familyHN 252 pts · 138 comments
Google calls Gemma 3 the most powerful AI model you can run on one GPU
theverge.com2025-03-20HN front pageabout the gemma 3 familyHN 127 pts · 100 comments
All coverage tracked for this model (26) — append-only, nothing is ever removed
TitleSourcePublishedFirst seenSignals
Open-jev: One-pass option scoring with Gemma 3 4B, similar to jevgithub.com2026-09-172026-10-07hn
Show HN: Gemma 3 inference in pure C++ with Metal accelerationgithub.com2026-07-042026-10-07hn
Selectively reducing eval awareness and murder in Gemma 3 27B via steeringlesswrong.com2026-03-122026-10-07hn
I have written gemma3 inference in pure Cgithub.com2026-01-262026-10-07hn
Own your AI: Learn how to fine-tune Gemma 3 270M and run it on-devicedevelopers.googleblog.com official2025-10-102026-10-07hn
Gemma 3 270M re-implemented in pure PyTorch for local tinkeringgithub.com2025-08-202026-10-07hn
Landrecords – cheap nationwide parcel dataset standardized using gemma3landrecords.us2025-08-202026-10-07hn
Gemma 3 270M: Compact model for hyper-efficient AIdevelopers.googleblog.com official2025-08-142026-10-07hn
I Gave Up Trying to Run Gemma 3 27B on AWS P3dn.24xlargemedium.com2025-08-112026-10-07hn
Gemma 3n: The Developer Guidesimonwillison.net2025-06-272026-10-07hn
Gemma 3n available in the open-source ecosystemhuggingface.co2025-06-262026-10-07hn
Introducing Gemma 3ndevelopers.googleblog.com official2025-06-262026-10-07hn
A C++ library to efficiently run Gemma-3N across various platformgithub.com2025-06-112026-10-07hn
Gemma 3n Architectural Innovations – Speculation and poking around in the modelold.reddit.com2025-05-252026-10-07hn
Gemma 3n preview: Mobile-first AIdevelopers.googleblog.com official2025-05-202026-10-07hn
Gemma 3 QAT Modelssimonwillison.net2025-04-192026-10-07hn
QAT optimized Gemma 3 slash VRAM needs 54GB –> 14.1GB while maintaining qualityold.reddit.com2025-04-192026-10-07hn
Gemma 3 QAT Models: Bringing AI to Consumer GPUsdevelopers.googleblog.com official2025-04-182026-10-07hn
Gemma3 Function Callingai.google.dev official2025-03-232026-10-07hn
Google calls Gemma 3 the most powerful AI model you can run on one GPUtheverge.com2025-03-202026-10-07hn
Fine-tune Google's Gemma 3unsloth.ai2025-03-192026-10-07hn
Google's Gemma 3 is an open source, multimodal single-GPU AI 128K context windowarstechnica.com2025-03-132026-10-07hn
Gemma3 – The current strongest model that fits on a single GPUollama.com2025-03-122026-10-07hn
Gemma 3: The most capable model you can run on a single GPU or TPUblog.google official2025-03-122026-10-07hn
Gemma 3: Google's new multimodal modelsai.google.dev official2025-03-122026-10-07hn
Gemma 3 Technical Report [pdf]storage.googleapis.com2025-03-122026-10-07hn

Papers

8 papers

Papers 8

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

  • Llama-3.2-1B-Instruct · 1.24B
    same size band (S), same task (text-generation), more providers
  • TinyLlama-1.1B-Chat-v1.0 · 1.10B
    same size band (S), same task (text-generation), permissive license
  • MiniCPM5-1B · 1.08B
    same size band (S), same task (text-generation), permissive license

Sources

3 sources

Sources

Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:38 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:09 UTC · Last verified 2026-10-25

Follow gemma-3-1b-pt

Watch this model

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