Best Hugging Face Models: The Complete Ranked List

TL;DR: This best Hugging Face models list ranks 300 models on real Hub data rather than opinion. By downloads the leader is sentence-transformers/all-MiniLM-L6-v2 at 248,935,735 in 30 days, which is 28.3% of every download across the whole list. By our composite quality score the leader is amazon/chronos-2 at 98/100. The finding worth carrying away: small embedding models, not chat models, are what people actually download.

Last updated: 7 August 2026. Data pulled: 7 August 2026, from the free public Hugging Face Hub API. Models ranked: 300. Downloads in the last 30 days: 878,230,936. All-time downloads across the list: 7,461,254,367.

Every “best Hugging Face models” article I can find ranks models by vibes, or by whichever LLM was trending the week it was written. This one ranks them by what the Hub itself reports: downloads, likes, licence, and how recently each model was touched.

I have spent 15 years in software and SEO and tested more than 500 AI tools, and the reason I keep coming back to download data is that it is the only honest signal of production use. Benchmarks tell you what a model can do. Downloads tell you what people are actually shipping. On Hugging Face those two answers are further apart than almost anyone expects.

The Numbers Worth Quoting

• 300 Hugging Face models ranked as of 7 August 2026, with 878,230,936 downloads in 30 days and 7.46B all-time, from 105 publishers.
• sentence-transformers/all-MiniLM-L6-v2 alone is 28.3% of 30-day downloads (248,935,735). It is a sentence-similarity model, not a chat model.
• Embeddings & Retrieval has only 25 models but takes 55.5% of downloads. Vision & Multimodal has 118 models and takes 19.9%.
• That is 19.5M downloads per model versus 1.5M, roughly a 13x difference.
• The top ten models take 56.7% of all downloads. The remaining 290 share 43.3%.
• 124 of 300 models (41%) are quantized or format-converted re-uploads of someone else’s weights, accounting for 15.0% of downloads.
• 80% are Apache-2.0 or MIT. 7 state no licence at all, five are non-commercial, and one is access-gated.
• Likes measure attention, not use. The most-liked model, moonshotai/Kimi-K3 at 10,248 likes, ranks only #89 by downloads.
• Maintenance is strong: median 28 days since the last update, 165 models updated within 30 days, and only one untouched for over a year.
• Quality scores land in a narrow band (78 to 98, median 83) because a model has to clear an adoption floor before it is ranked at all.

Every figure is reproducible from the tables below. Methodology and citation line at the bottom.

What This Best Hugging Face Models List Covers

The Hugging Face Hub hosts well over a million models. Ranking all of them would be meaningless, because most have never been downloaded by anyone but their author. So inclusion here is automatic and has a floor: a model is ranked when it clears a minimum adoption threshold, at least 1,000 downloads in the last 30 days or at least 50 likes. That leaves the 300 models in this report.

Two consequences worth stating up front.

First, this is a list of what is being used, not what is newest or most impressive on a benchmark. A model released last week will not appear until people download it.

Second, because everything here already cleared the floor, the quality scores cluster tightly. The range is 78 to 98 with a median of 83. The score is useful for ordering models inside a task category; it is not a verdict on whether a model is good, and I would not read a 83 as a warning.

No model pays to be listed or ranked. There are no affiliate links in this report, because Hugging Face models are free.

Which Kinds of Model Are Actually Downloaded Most?

Embeddings & Retrieval models. Not chat models, not image generators. 25 embedding and retrieval models take 55.5% of every download in this list, while 118 Vision & Multimodal models take 19.9%.

Task group	Models	30-day downloads	Share of downloads	Downloads per model
Embeddings & Retrieval	25	487,105,326	55.5%	19.5M
Vision & Multimodal	118	174,618,155	19.9%	1.5M
Text Generation & Chat	72	99,263,688	11.3%	1.4M
Tabular, Time-Series & Robotics	4	42,004,898	4.8%	10.5M
Other	43	37,477,369	4.3%	871.6k
Speech & Audio	26	32,042,609	3.6%	1.2M
Text Classification	10	4,931,215	0.6%	493.1k
Image Generation	2	787,676	0.1%	393.8k
All 8 groups	300	878,230,936	100%	2.9M

This is the single most useful thing in the dataset, and it is almost the exact inverse of the public conversation.

The reason is structural. An embedding model is infrastructure: if you run semantic search, a RAG pipeline, or a recommendation system, you call it on every document and every query, and you pull the weights into every container you deploy. A chat model is a destination: you download it once, or you never download it at all because you call somebody’s API instead.

So download counts on Hugging Face measure something specific. They measure how often a model gets pulled into a build, which correlates with infrastructure use rather than user-facing excitement. That is worth knowing before you cite a download number as evidence that a model is “the best”.

The practical read for a buyer: if you are choosing an embedding model, this data is extremely informative, because the whole market is here and heavily used. If you are choosing a chat model, treat downloads as one weak signal among several.

How Concentrated Are Hugging Face Downloads?

Extremely. sentence-transformers/all-MiniLM-L6-v2 alone accounts for 28.3% of 30-day downloads. The top three take 44.6%, the top ten 56.7%, and the top fifty 80.1%.

The median model in this list gets 576,071 downloads in 30 days. So the gap between the leader and the middle is roughly four orders of magnitude.

What makes the leader interesting is how old and how small it is. `all-MiniLM-L6-v2` is a compact sentence-transformer that has been the default embedding model in countless tutorials, frameworks and starter templates for years. Its dominance is a lesson in defaults: being the thing that gets copy-pasted into every quickstart compounds far faster than being the best model in a benchmark table.

The 10 Best Hugging Face Models by Quality Score

Rank	Model	Task	30-day downloads	Likes	Licence	Quality score
1	amazon/chronos-2	time-series-forecasting	31,142,302	393	apache-2.0	98/100
2	google/gemma-4-E2B-it	any-to-any	4,002,947	874	apache-2.0	97/100
3	ibm-research/MoLFormer-XL-both-10pct	feature-extraction	265,114	36	apache-2.0	96/100
4	autogluon/chronos-2	time-series-forecasting	10,712,794	48	apache-2.0	95/100
5	nvidia/Qwen3.6-35B-A3B-NVFP4	text-generation	10,696,241	539	apache-2.0	95/100
6	unsloth/Qwen3.6-27B-NVFP4	image-text-to-text	3,635,371	267	apache-2.0	95/100
7	cyankiwi/Qwen3.6-27B-AWQ-INT4	image-text-to-text	2,515,205	102	apache-2.0	95/100
8	baidu/Unlimited-OCR	image-text-to-text	2,836,694	3,941	mit	94/100
9	google/gemma-4-31B-it-qat-w4a16-ct	image-text-to-text	2,238,004	54	apache-2.0	94/100
10	nvidia/parakeet-tdt-0.6b-v2	automatic-speech-recognition	671,034	1,532	cc-by-4.0	94/100

This ranking weighs adoption alongside growth, maintenance and trust, so it surfaces models that are both used and well kept rather than only enormous.

amazon/chronos-2 leads at 98/100 on 31,142,302 downloads. Note what happens next in the table: several entries have download counts far below the raw leaders, and still place highly, because they are growing fast, freshly updated, properly licensed and not gated.

The 10 Most-Downloaded Hugging Face Models

Rank	Model	Task	30-day downloads	Share	Quality score
1	sentence-transformers/all-MiniLM-L6-v2	sentence-similarity	248,935,735	28.3%	89/100
2	cross-encoder/ms-marco-MiniLM-L6-v2	text-ranking	85,315,799	9.7%	78/100
3	sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2	sentence-similarity	57,173,935	6.5%	82/100
4	amazon/chronos-2	time-series-forecasting	31,142,302	3.5%	98/100
5	nomic-ai/nomic-embed-text-v1.5	sentence-similarity	15,099,753	1.7%	85/100
6	intfloat/multilingual-e5-small	sentence-similarity	14,804,481	1.7%	84/100
7	Qwen/Qwen3.5-9B	image-text-to-text	12,159,684	1.4%	93/100
8	sentence-transformers/paraphrase-multilingual-mpnet-base-v2	sentence-similarity	11,111,585	1.3%	78/100
9	google/gemma-4-31B-it	image-text-to-text	11,028,273	1.3%	88/100
10	google/gemma-4-26B-A4B-it	image-text-to-text	10,976,594	1.2%	85/100

Compare the two tables. They share almost nothing, and that is the point.

sentence-transformers/all-MiniLM-L6-v2 is first here and does not lead on quality score. amazon/chronos-2 leads on quality score and is not first here. Neither ranking is wrong: downloads tell you what is battle-tested and safe to standardise on, while the quality score tells you what is well maintained and moving. For infrastructure you will keep for years, weight downloads more heavily. For something you are adopting now, weight the score.

How Is the Quality Score Calculated?

Four weighted pillars: adoption 40%, maintenance 25%, growth 20%, trust 15%.

What each one means:

• Adoption (40%) blends all-time downloads and likes, log-scaled so a handful of enormous outliers do not flatten everything below them.
• Maintenance (25%) is how recently the model was updated, decaying the longer it sits untouched.
• Growth (20%) is the 30-day change in downloads, measured against our own snapshot from roughly a month earlier. It is blank for models too new to have history, and those are scored neutrally rather than penalised.
• Trust (15%) is structural rather than subjective: does the model state a licence, is its author identified, and is it free of an access gate. It is the pillar that most often separates an otherwise-identical pair.

Hugging Face also exposes an internal “trending score”, and 201 of the 300 models carry a non-zero one. Its scale is not publicly documented, so it is not used in the ranking at all.

Are Hugging Face Models Free to Use? Licences in Practice

Mostly yes, and more permissively than people assume. 198 models are Apache-2.0 and 41 are MIT, so 80% of the list carries a licence that allows commercial use with minimal conditions.

License	Models	Share
apache-2.0	198	66%
mit	41	14%
other	36	12%
none stated	7	2%
cc-by-4.0	6	2%
cc-by-nc-4.0	4	1%
openrail	2	1%
openmdw-1.1	2	1%

The exceptions are where the care is needed. Seven models state no licence at all, which legally is the most restrictive outcome rather than the least, because absent a grant you have no permission. five carry a non-commercial licence, and one is access-gated, meaning you must accept terms before downloading.

A caveat on the “other” bucket: Hugging Face reports it for custom licences, which includes several widely used model families whose terms are real but non-standard. If a model matters to your product, read its actual licence file rather than trusting the tag.

Which Hugging Face Models Are Growing Fastest?

unsloth/Qwen3.6-35B-A3B-NVFP4 grew 488% over the last 30 days, followed by zai-org/GLM-5.2 at 396%. Of the 234 models with enough history to measure, 147 grew and 85 declined.

Model	30-day trend	30-day downloads	Task	Quality score
unsloth/Qwen3.6-35B-A3B-NVFP4	+488%	1,998,933	image-text-to-text	93/100
zai-org/GLM-5.2	+396%	2,430,330	text-generation	92/100
Qwen/Qwen3.5-122B-A10B	+252%	2,138,802	image-text-to-text	88/100
DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF	+250%	2,217,339	image-text-to-text	93/100
nvidia/nemotron-3.5-asr-streaming-0.6b	+234%	1,052,774	automatic-speech-recognition	92/100
ibm-granite/granite-4.1-8b	+212%	4,213,980	text-generation	87/100
deepseek-ai/DeepSeek-V4-Flash-0731	+198%	702,709	text-generation	92/100
datalab-to/surya-ocr-2	+195%	1,235,692	image-text-to-text	90/100
wikeeyang/Flux2-Klein-9B-True-V2	+186%	787,676	text-to-image	91/100
Qwen/Qwen3.5-0.8B-Base	+172%	846,908	image-text-to-text	85/100
cyankiwi/Qwen3-VL-8B-Instruct-AWQ-4bit	+153%	405,196	image-text-to-text	91/100
cyankiwi/GLM-4.7-Flash-AWQ-4bit	+143%	498,198	text-generation	93/100

Two patterns in that list. Most risers are new open-weight releases finding their audience in their first weeks. A minority are quantized or format-converted versions of a model that already existed, which is a different phenomenon: the weights are not new, the packaging is.

That second pattern is worth its own number, because it is much bigger across the list than it is at the top of the growth table. 124 of the 300 ranked models (41%) are quantized or converted re-uploads in formats like GGUF, AWQ, NVFP4 and INT4, together 15.0% of downloads. Whole publishers on this list exist to do nothing else. It tells you where the real bottleneck is: not capability, but getting existing capability onto hardware people can afford.

And the other direction:

Model	30-day trend	30-day downloads	Task	Quality score
cyankiwi/gemma-4-26B-A4B-it-AWQ-4bit	-40%	3,331,398	image-text-to-text	78/100
tabularisai/multilingual-sentiment-analysis	-31%	329,648	text-classification	78/100
cyankiwi/Qwen3.5-9B-AWQ-4bit	-27%	287,482	image-text-to-text	78/100
lmstudio-community/gemma-4-26B-A4B-it-QAT-MLX-4bit	-23%	844,724	image-text-to-text	78/100
google/gemma-4-26B-A4B-it	-23%	10,976,594	image-text-to-text	85/100
answerdotai/answerai-colbert-small-v1	-23%	250,948	other	78/100

cyankiwi/gemma-4-26B-A4B-it-AWQ-4bit is down 40%, the steepest decline here. Most decliners are the previous generation of the same model family, which is what healthy replacement looks like in download data rather than a sign of a problem.

Do Likes Tell You Anything?

Not about usage. Likes measure attention, and the gap between attention and use on this list is large enough to be worth a section.

Model	Likes	30-day downloads	Rank by downloads	Task
moonshotai/Kimi-K3	10,248	1,308,186	#89	image-text-to-text
deepseek-ai/DeepSeek-V4-Pro	5,392	1,561,291	#84	text-generation
sentence-transformers/all-MiniLM-L6-v2	5,186	248,935,735	#1	sentence-similarity
bigscience/bloom	5,033	0	#272	text-generation
openai/gpt-oss-20b	4,884	8,229,095	#16	text-generation
zai-org/GLM-5.2	4,882	2,430,330	#58	text-generation
baidu/Unlimited-OCR	3,941	2,836,694	#45	image-text-to-text
google/gemma-4-31B-it	3,469	11,028,273	#9	image-text-to-text
Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled	2,929	0	#280	image-text-to-text
MiniMaxAI/MiniMax-H3	2,861	0	#275	image-text-to-video

moonshotai/Kimi-K3 is the most-liked model at 10,248 likes, and ranks #89 by downloads with 1,308,186 in 30 days. The list has 171,360 likes in total, and they cluster on frontier-scale releases that most people admire and few people can actually run.

That is not a criticism of likes. It just means they answer “what is exciting” while downloads answer “what is deployed”. Both are in the tables so you can read either.

Are These Models Maintained?

Better than any of the extension or plugin ecosystems we track. The median model here was updated 28 days ago, 165 of 300 (55%) within 30 days, 243 within 90, and only one model has gone more than a year, the oldest at 377 days.

There is a caveat about what “updated” means on the Hub. A model’s last-modified date changes when anything in the repository changes, which includes a README edit or a new quantization variant, not only new weights. So read it as a signal that somebody is still paying attention to the repo, not that the model itself was retrained.

For comparison, in our AI Chrome extensions report fifteen extensions with real user bases had gone over a year without an update. Model publishing has a healthier maintenance culture, probably because a model repo is a research artefact with a reputation attached.

Best Hugging Face Models by Task

Nobody needs “the best model”. They need an embedding model, or an OCR model, or something that transcribes audio. Each group below is the top six by quality score within that task, with downloads, likes and licence alongside.

Best embedding and retrieval models

These turn text into vectors for semantic search, RAG pipelines, and re-ranking. They are small, unglamorous, and by download volume they are the most used models on the entire Hub.

25 models in this group, 487,105,326 downloads in the last 30 days (55.5% of the total).

Model	30-day downloads	Likes	Licence	Quality score
ibm-research/MoLFormer-XL-both-10pct	265,114	36	apache-2.0	96/100
zeroentropy/zerank-2-reranker	358,636	109	apache-2.0	91/100
nvidia/Nemotron-3-Embed-1B-BF16	467,453	129	other	90/100
sentence-transformers/all-MiniLM-L6-v2	248,935,735	5,186	apache-2.0	89/100
Qwen/Qwen3-VL-Embedding-8B	2,206,090	468	apache-2.0	89/100
nvidia/llama-nemotron-embed-1b-v2	810,123	61	other	88/100

Best vision and multimodal models

Vision-language models that handle image understanding, visual question answering, OCR, and multimodal chat in a single model. This is the most crowded category on the list.

118 models in this group, 174,618,155 downloads in the last 30 days (19.9% of the total).

Model	30-day downloads	Likes	Licence	Quality score
google/gemma-4-E2B-it	4,002,947	874	apache-2.0	97/100
unsloth/Qwen3.6-27B-NVFP4	3,635,371	267	apache-2.0	95/100
cyankiwi/Qwen3.6-27B-AWQ-INT4	2,515,205	102	apache-2.0	95/100
baidu/Unlimited-OCR	2,836,694	3,941	mit	94/100
google/gemma-4-31B-it-qat-w4a16-ct	2,238,004	54	apache-2.0	94/100
Qwen/Qwen3.5-9B	12,159,684	1,794	apache-2.0	93/100

Best text generation and chat models

The models behind most chat assistants and text-completion tools built on Hugging Face, including the open-weight releases people actually deploy rather than only benchmark.

72 models in this group, 99,263,688 downloads in the last 30 days (11.3% of the total).

Model	30-day downloads	Likes	Licence	Quality score
nvidia/Qwen3.6-35B-A3B-NVFP4	10,696,241	539	apache-2.0	95/100
farbodtavakkoli/OTel-2.0-LLM-31B-IT	3,688,285	8	apache-2.0	93/100
prism-ml/Bonsai-27B-gguf	2,650,023	744	apache-2.0	93/100
cyankiwi/Qwen3-Coder-30B-A3B-Instruct-AWQ-4bit	615,127	56	apache-2.0	93/100
LiquidAI/LFM2.5-1.2B-Instruct	555,431	647	other	93/100
cyankiwi/GLM-4.7-Flash-AWQ-4bit	498,198	56	mit	93/100

Best tabular, time-series and robotics models

An emerging mix of forecasting and robotics foundation models. Tiny by model count and punching far above its weight on downloads.

4 models in this group, 42,004,898 downloads in the last 30 days (4.8% of the total).

Model	30-day downloads	Likes	Licence	Quality score
amazon/chronos-2	31,142,302	393	apache-2.0	98/100
autogluon/chronos-2	10,712,794	48	apache-2.0	95/100
Datadog/Toto-Open-Base-1.0	149,802	142	apache-2.0	84/100
nvidia/Alpamayo-R1-10B	0	427	openmdw-1.1	79/100

Other notable models

Models whose Hugging Face task tag does not map cleanly onto the groups above, including reinforcement learning, depth estimation, and unclassified releases.

43 models in this group, 37,477,369 downloads in the last 30 days (4.3% of the total).

Model	30-day downloads	Likes	Licence	Quality score
Comfy-Org/Wan_2.2_ComfyUI_Repackaged	5,483,689	828	none	92/100
Comfy-Org/Qwen-Image-Edit_ComfyUI	873,011	453	apache-2.0	92/100
kernels-community/flash-attn3	470,536	48	bsd-3-clause	91/100
biohub/ESMFold2	462,324	50	mit	91/100
fastino/gliner2-large-v1	449,916	91	apache-2.0	90/100
nvidia/Cosmos3-Nano	313,164	331	other	90/100

Best speech and audio models

Transcription, text-to-speech, and audio classification. Speech recognition in particular has quietly become one of the most reliably downloaded categories on the Hub.

26 models in this group, 32,042,609 downloads in the last 30 days (3.6% of the total).

Model	30-day downloads	Likes	Licence	Quality score
nvidia/parakeet-tdt-0.6b-v2	671,034	1,532	cc-by-4.0	94/100
nvidia/nemotron-3.5-asr-streaming-0.6b	1,052,774	1,002	other	92/100
handy-computer/nemotron-3.5-asr-streaming-0.6b-gguf	1,945,949	3	other	91/100
handy-computer/parakeet-unified-en-0.6b-gguf	1,787,790	3	cc-by-4.0	91/100
distil-whisper/distil-large-v3	1,612,178	378	mit	91/100
pnnbao-ump/VieNeu-TTS-v3-Turbo	343,303	47	apache-2.0	90/100

Best text classification models

Sentiment, masked-token prediction, named-entity tagging, and safety classification. Old-fashioned by 2026 standards and still doing enormous amounts of production work.

10 models in this group, 4,931,215 downloads in the last 30 days (0.6% of the total).

Model	30-day downloads	Likes	Licence	Quality score
protectai/deberta-v3-base-prompt-injection-v2	293,535	112	apache-2.0	93/100
vinai/phobert-base	162,285	71	mit	87/100
biohub/ESMC-6B	2,059,684	24	mit	85/100
openai/privacy-filter	516,042	1,718	apache-2.0	85/100
protectai/unbiased-toxic-roberta-onnx	159,849	7	apache-2.0	85/100
SamLowe/roberta-base-go_emotions	904,023	685	mit	84/100

Best image generation models

Text-to-image generation. A small category here by count, because most image-generation traffic sits with a handful of well-known checkpoints and their community fine-tunes.

2 models in this group, 787,676 downloads in the last 30 days (0.1% of the total).

Model	30-day downloads	Likes	Licence	Quality score
wikeeyang/Flux2-Klein-9B-True-V2	787,676	182	other	91/100
LoliRimuru/moeFussion	0	294	creativeml-openrail-m	79/100

The Full List: All 300 Hugging Face Models Ranked

Every ranked model, ordered by quality score, with the inputs shown so you can check the arithmetic. Downloads are as reported by the Hub API. Click any model to open its official Hugging Face page and model card.

#	Model	Publisher	Task	30-day downloads	All-time	Likes	Updated	Licence	Quality
1	amazon/chronos-2	amazon	time-series-forecasting	31,142,302	134,067,414	393	2026-06-05	apache-2.0	98
2	google/gemma-4-E2B-it	google	any-to-any	4,002,947	13,707,085	874	2026-07-20	apache-2.0	97
3	ibm-research/MoLFormer-XL-both-10pct	ibm-research	feature-extraction	265,114	8,820,616	36	2026-07-23	apache-2.0	96
4	autogluon/chronos-2	autogluon	time-series-forecasting	10,712,794	59,928,012	48	2026-06-05	apache-2.0	95
5	nvidia/Qwen3.6-35B-A3B-NVFP4	nvidia	text-generation	10,696,241	20,355,559	539	2026-06-12	apache-2.0	95
6	unsloth/Qwen3.6-27B-NVFP4	unsloth	image-text-to-text	3,635,371	5,738,175	267	2026-07-12	apache-2.0	95
7	cyankiwi/Qwen3.6-27B-AWQ-INT4	cyankiwi	image-text-to-text	2,515,205	6,480,211	102	2026-07-21	apache-2.0	95
8	baidu/Unlimited-OCR	baidu	image-text-to-text	2,836,694	4,073,663	3,941	2026-07-29	mit	94
9	google/gemma-4-31B-it-qat-w4a16-ct	google	image-text-to-text	2,238,004	3,569,149	54	2026-07-20	apache-2.0	94
10	nvidia/parakeet-tdt-0.6b-v2	nvidia	automatic-speech-recognition	671,034	11,258,343	1,532	2026-06-29	cc-by-4.0	94
11	Qwen/Qwen3.5-9B	Qwen	image-text-to-text	12,159,684	46,017,566	1,794	2026-03-02	apache-2.0	93
12	farbodtavakkoli/OTel-2.0-LLM-31B-IT	farbodtavakkoli	text-generation	3,688,285	3,688,285	8	2026-08-03	apache-2.0	93
13	datalab-to/chandra-ocr-2	datalab-to	image-text-to-text	2,891,159	7,465,965	468	2026-06-26	openrail	93
14	prism-ml/Bonsai-27B-gguf	prism-ml	text-generation	2,650,023	2,650,029	744	2026-07-17	apache-2.0	93
15	DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF	DavidAU	image-text-to-text	2,217,339	2,217,339	1,675	2026-08-05	apache-2.0	93
16	unsloth/Qwen3.6-35B-A3B-NVFP4	unsloth	image-text-to-text	1,998,933	2,449,781	107	2026-07-12	apache-2.0	93
17	PaddlePaddle/PP-DocLayoutV3_safetensors	PaddlePaddle	object-detection	789,663	2,404,543	38	2026-07-08	apache-2.0	93
18	cyankiwi/Qwen3-Coder-30B-A3B-Instruct-AWQ-4bit	cyankiwi	text-generation	615,127	2,574,356	56	2026-07-21	apache-2.0	93
19	LiquidAI/LFM2.5-1.2B-Instruct	LiquidAI	text-generation	555,431	2,133,046	647	2026-08-05	other	93
20	cyankiwi/GLM-4.7-Flash-AWQ-4bit	cyankiwi	text-generation	498,198	2,302,932	56	2026-07-21	mit	93
21	protectai/deberta-v3-base-prompt-injection-v2	protectai	text-classification	293,535	5,914,864	112	2026-07-09	apache-2.0	93
22	Comfy-Org/Wan_2.2_ComfyUI_Repackaged	Comfy-Org	other	5,483,689	77,389,401	828	2026-07-03	none	92
23	zai-org/GLM-OCR	zai-org	image-text-to-text	3,591,079	28,293,943	1,971	2026-05-19	mit	92
24	zai-org/GLM-5.2	zai-org	text-generation	2,430,330	2,790,370	4,882	2026-07-02	mit	92
25	nvidia/nemotron-3.5-asr-streaming-0.6b	nvidia	automatic-speech-recognition	1,052,774	1,285,976	1,002	2026-08-05	other	92
26	Comfy-Org/Qwen-Image-Edit_ComfyUI	Comfy-Org	other	873,011	9,415,300	453	2026-07-01	apache-2.0	92
27	openbmb/MiniCPM-o-4_5	openbmb	any-to-any	819,702	1,712,787	1,458	2026-08-03	apache-2.0	92
28	deepseek-ai/DeepSeek-V4-Flash-0731	deepseek-ai	text-generation	702,709	702,709	2,709	2026-08-01	mit	92
29	LilaRest/gemma-4-31B-it-NVFP4-turbo	LilaRest	text-generation	633,897	1,560,465	303	2026-07-18	apache-2.0	92
30	unsloth/gemma-4-E4B-it-unsloth-bnb-4bit	unsloth	image-text-to-text	555,702	1,587,292	22	2026-07-17	apache-2.0	92
31	Qwen/Qwen3.6-27B-FP8	Qwen	image-text-to-text	7,757,906	24,051,224	342	2026-04-24	apache-2.0	91
32	Qwen/Qwen3.6-27B	Qwen	image-text-to-text	6,779,079	20,025,407	2,192	2026-04-24	apache-2.0	91
33	nvidia/Gemma-4-31B-IT-NVFP4	nvidia	text-generation	2,582,121	9,518,975	548	2026-07-13	other	91
34	handy-computer/nemotron-3.5-asr-streaming-0.6b-gguf	handy-computer	automatic-speech-recognition	1,945,949	2,488,982	3	2026-06-29	other	91
35	handy-computer/parakeet-unified-en-0.6b-gguf	handy-computer	automatic-speech-recognition	1,787,790	2,258,246	3	2026-06-28	cc-by-4.0	91
36	nvidia/Qwen3.6-27B-NVFP4	nvidia	text-generation	1,746,168	2,486,736	420	2026-06-30	apache-2.0	91
37	nvidia/GLM-5.2-NVFP4	nvidia	text-generation	1,678,603	2,294,144	306	2026-06-26	mit	91
38	distil-whisper/distil-large-v3	distil-whisper	automatic-speech-recognition	1,612,178	19,509,097	378	2026-04-21	mit	91
39	h2oai/h2ovl-mississippi-2b	h2oai	text-generation	1,207,565	9,183,037	43	2026-07-16	apache-2.0	91
40	wikeeyang/Flux2-Klein-9B-True-V2	wikeeyang	text-to-image	787,676	1,100,345	182	2026-07-13	other	91
41	prism-ml/Ternary-Bonsai-27B-gguf	prism-ml	text-generation	784,092	784,100	1,171	2026-07-18	apache-2.0	91
42	google/gemma-4-12B-it-qat-q4_0-unquantized	google	any-to-any	565,821	856,412	72	2026-07-20	apache-2.0	91
43	google/gemma-4-E4B-it-qat-w4a16-ct	google	any-to-any	496,833	855,605	14	2026-07-20	apache-2.0	91
44	kernels-community/flash-attn3	kernels-community	other	470,536	2,467,518	48	2026-06-27	bsd-3-clause	91
45	biohub/ESMFold2	biohub	other	462,324	847,590	50	2026-07-28	mit	91
46	XiaomiMiMo/MiMo-V2.5	XiaomiMiMo	text-generation	405,348	952,764	389	2026-07-09	mit	91
47	cyankiwi/Qwen3-VL-8B-Instruct-AWQ-4bit	cyankiwi	image-text-to-text	405,196	1,043,333	16	2026-07-21	apache-2.0	91
48	zeroentropy/zerank-2-reranker	zeroentropy	text-ranking	358,636	790,121	109	2026-07-24	apache-2.0	91
49	cyankiwi/gemma-4-E4B-it-AWQ-INT4	cyankiwi	any-to-any	301,788	802,384	4	2026-07-21	apache-2.0	91
50	datalab-to/surya-ocr-2	datalab-to	image-text-to-text	1,235,692	1,721,444	90	2026-05-27	openrail	90
51	h2oai/h2ovl-mississippi-800m	h2oai	text-generation	1,217,640	8,905,522	40	2026-07-16	apache-2.0	90
52	openbmb/MiniCPM5-1B	openbmb	text-generation	927,647	1,437,587	1,037	2026-05-26	apache-2.0	90
53	prism-ml/Bonsai-27B-mlx-1bit	prism-ml	text-generation	664,420	664,436	204	2026-07-14	apache-2.0	90
54	maci0/Qwopus3.6-27B-Coder-NVFP4	maci0	image-text-to-text	557,539	573,223	3	2026-08-04	apache-2.0	90
55	empero-ai/Qwythos-9B-v2-GGUF	empero-ai	image-text-to-text	502,950	502,950	235	2026-07-12	apache-2.0	90
56	nvidia/Nemotron-3-Embed-1B-BF16	nvidia	sentence-similarity	467,453	470,737	129	2026-08-06	other	90
57	fastino/gliner2-large-v1	fastino	other	449,916	1,784,107	91	2026-05-19	apache-2.0	90
58	allenai/Olmo-3-7B-Instruct	allenai	text-generation	429,485	2,244,582	141	2026-06-25	apache-2.0	90
59	pnnbao-ump/VieNeu-TTS-v3-Turbo	pnnbao-ump	text-to-speech	343,303	561,054	47	2026-07-11	apache-2.0	90
60	nvidia/Cosmos3-Nano	nvidia	other	313,164	576,455	331	2026-07-09	other	90
61	unsloth/gemma-4-E4B-it-qat-GGUF	unsloth	any-to-any	285,969	555,629	151	2026-07-17	apache-2.0	90
62	sentence-transformers/all-MiniLM-L6-v2	sentence-transformers	sentence-similarity	248,935,735	3,515,354,383	5,186	2026-06-01	apache-2.0	89
63	google/gemma-4-12B-it	google	any-to-any	2,963,990	6,655,237	1,407	2026-07-20	apache-2.0	89
64	nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4	nvidia	text-generation	2,742,676	8,491,344	415	2026-05-01	other	89
65	Qwen/Qwen3-VL-Embedding-8B	Qwen	sentence-similarity	2,206,090	8,310,725	468	2026-04-16	apache-2.0	89
66	deepseek-ai/DeepSeek-V4-Pro	deepseek-ai	text-generation	1,561,291	9,594,877	5,392	2026-06-22	mit	89
67	Bahushruth/Qwen3.6-35B-A3B-abliterated-v4	Bahushruth	text-generation	980,869	982,849	6	2026-07-03	apache-2.0	89
68	handy-computer/cohere-transcribe-03-2026-gguf	handy-computer	automatic-speech-recognition	957,514	1,212,680	3	2026-06-28	apache-2.0	89
69	cyankiwi/gemma-4-12B-it-AWQ-INT4	cyankiwi	any-to-any	523,535	1,054,032	8	2026-07-21	apache-2.0	89
70	Jackrong/Qwopus3.6-35B-A3B-Coder-MTP-GGUF	Jackrong	image-text-to-text	479,958	739,725	222	2026-07-09	apache-2.0	89
71	google/gemma-4-31B-it	google	image-text-to-text	11,028,273	45,741,153	3,469	2026-07-20	apache-2.0	88
72	Qwen/Qwen3.6-35B-A3B-FP8	Qwen	image-text-to-text	8,821,882	25,420,039	341	2026-04-24	apache-2.0	88
73	google/gemma-4-E4B-it	google	any-to-any	5,295,850	24,602,508	1,454	2026-07-20	apache-2.0	88
74	Comfy-Org/z_image_turbo	Comfy-Org	other	5,138,962	28,100,403	802	2026-07-02	none	88
75	Qwen/Qwen3.5-2B	Qwen	image-text-to-text	2,636,097	10,436,224	351	2026-03-02	apache-2.0	88
76	Qwen/Qwen3.5-122B-A10B	Qwen	image-text-to-text	2,138,802	6,094,611	605	2026-04-24	apache-2.0	88
77	nvidia/Kimi-K2.7-Code-NVFP4	nvidia	text-generation	915,103	917,668	9	2026-07-06	other	88
78	nvidia/llama-nemotron-embed-1b-v2	nvidia	feature-extraction	810,123	3,351,540	61	2026-05-20	other	88
79	handy-computer/parakeet-tdt-0.6b-v3-gguf	handy-computer	automatic-speech-recognition	604,725	730,348	1	2026-06-28	cc-by-4.0	88
80	deepseek-ai/DeepSeek-V4-Flash-DSpark	deepseek-ai	text-generation	564,628	681,941	247	2026-07-04	mit	88
81	nvidia/MiniMax-M3-NVFP4	nvidia	text-generation	542,672	662,597	75	2026-06-26	other	88
82	handy-computer/whisper-medium-gguf	handy-computer	automatic-speech-recognition	503,504	621,785	0	2026-06-28	apache-2.0	88
83	AngelSlim/Hy3-GGUF	AngelSlim	text-generation	446,433	446,433	179	2026-07-21	apache-2.0	88
84	poolside/Laguna-S-2.1-NVFP4	poolside	text-generation	425,357	425,425	176	2026-08-01	openmdw-1.1	88
85	nvidia/Nemotron-Labs-Diffusion-8B-Base	nvidia	text-generation	413,039	2,880,672	7	2026-06-03	other	88
86	MongoDB/mdbr-leaf-ir	MongoDB	sentence-similarity	384,372	534,937	66	2026-07-20	apache-2.0	88
87	palmfuture/Qwen3.6-35B-A3B-GPTQ-Int4	palmfuture	image-text-to-text	239,755	989,341	29	2026-07-05	apache-2.0	88
88	ibm-granite/granite-4.1-8b	ibm-granite	text-generation	4,213,980	5,108,473	246	2026-05-04	apache-2.0	87
89	Qwen/Qwen3-Reranker-4B	Qwen	text-ranking	2,720,660	9,060,791	151	2026-04-16	apache-2.0	87
90	google/diffusiongemma-26B-A4B-it	google	image-text-to-text	1,952,156	3,911,325	1,162	2026-07-15	apache-2.0	87
91	jhgan/ko-sroberta-multitask	jhgan	sentence-similarity	1,733,081	22,549,162	150	2026-06-16	none	87
92	nm-testing/SmolLM-1.7B-Instruct-quantized.w4a16	nm-testing	text-generation	1,431,131	4,399,463	0	2026-07-22	apache-2.0	87
93	unsloth/gemma-4-26B-A4B-it-GGUF	unsloth	image-text-to-text	1,329,673	9,642,752	1,034	2026-07-17	apache-2.0	87
94	moonshotai/Kimi-K3	moonshotai	image-text-to-text	1,308,186	1,308,242	10,248	2026-07-27	other	87
95	nvidia/DeepSeek-V4-Flash-NVFP4	nvidia	text-generation	957,168	1,706,813	94	2026-06-15	mit	87
96	Lorbus/Qwen3.6-27B-int4-AutoRound	Lorbus	image-text-to-text	885,082	3,772,759	130	2026-04-22	apache-2.0	87
97	nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16	nvidia	text-generation	876,869	8,290,151	806	2026-07-23	other	87
98	unsloth/gemma-4-E4B-it-GGUF	unsloth	image-text-to-text	535,556	4,518,808	578	2026-07-17	apache-2.0	87
99	unsloth/gemma-4-31B-it-GGUF	unsloth	image-text-to-text	516,606	4,249,503	564	2026-07-17	apache-2.0	87
100	sakamakismile/Ornith-1.0-35B-NVFP4	sakamakismile	image-text-to-text	472,603	577,837	25	2026-06-25	mit	87
101	google/gemma-4-12B-it-qat-q4_0-gguf	google	any-to-any	302,630	891,422	264	2026-07-17	apache-2.0	87
102	vinai/phobert-base	vinai	fill-mask	162,285	12,039,782	71	2026-08-03	mit	87
103	nomic-ai/nomic-embed-text-v1	nomic-ai	sentence-similarity	4,848,481	51,948,593	580	2026-04-07	apache-2.0	86
104	mistralai/Voxtral-Mini-4B-Realtime-2602	mistralai	automatic-speech-recognition	2,219,534	8,266,288	936	2026-03-11	apache-2.0	86
105	RedHatAI/Qwen3.6-35B-A3B-NVFP4	RedHatAI	other	1,865,548	8,524,326	168	2026-07-13	apache-2.0	86
106	nvidia/parakeet-ctc-1.1b	nvidia	automatic-speech-recognition	1,778,697	7,727,300	56	2026-08-05	cc-by-4.0	86
107	Kijai/LTX2.3_comfy	Kijai	other	1,036,272	6,314,131	573	2026-07-28	other	86
108	circlestone-labs/Anima	circlestone-labs	other	798,025	3,544,807	2,026	2026-07-24	other	86
109	Synaptics/yolo	Synaptics	other	470,310	1,616,708	0	2026-05-11	agpl-3.0	86
110	Lightricks/LTX-2	Lightricks	image-to-video	407,211	8,038,860	1,769	2026-08-04	other	86
111	nomic-ai/nomic-embed-text-v1.5	nomic-ai	sentence-similarity	15,099,753	133,000,287	886	2026-04-07	apache-2.0	85
112	google/gemma-4-26B-A4B-it	google	image-text-to-text	10,976,594	47,144,620	1,360	2026-07-20	apache-2.0	85
113	sentence-transformers/all-MiniLM-L12-v2	sentence-transformers	sentence-similarity	3,356,459	247,572,318	325	2026-03-31	apache-2.0	85
114	jinaai/jina-embeddings-v3	jinaai	feature-extraction	3,061,380	77,354,013	1,152	2026-04-08	cc-by-nc-4.0	85
115	biohub/ESMC-6B	biohub	fill-mask	2,059,684	4,310,718	24	2026-06-03	mit	85
116	Qwen/Qwen3.5-122B-A10B-FP8	Qwen	image-text-to-text	1,231,376	5,452,442	112	2026-04-24	apache-2.0	85
117	nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-FP8	nvidia	any-to-any	1,112,543	1,429,166	61	2026-05-05	other	85
118	litert-community/gemma-4-E2B-it-litert-lm	litert-community	other	1,088,424	3,846,225	386	2026-07-10	apache-2.0	85
119	iitolstykh/mivolo_v2	iitolstykh	other	904,861	23,268,094	32	2026-03-11	apache-2.0	85
120	Qwen/Qwen3.5-0.8B-Base	Qwen	image-text-to-text	846,908	1,719,432	91	2026-04-23	apache-2.0	85
121	sakamakismile/Qwen3.6-27B-Text-NVFP4-MTP	sakamakismile	text-generation	693,531	2,169,124	80	2026-04-29	apache-2.0	85
122	Qwen/Qwen3-VL-Reranker-2B	Qwen	text-ranking	580,453	2,162,488	212	2026-04-16	apache-2.0	85
123	Qwen/Qwen3.5-122B-A10B-GPTQ-Int4	Qwen	image-text-to-text	555,055	1,550,717	46	2026-04-24	apache-2.0	85
124	openai/privacy-filter	openai	token-classification	516,042	1,359,891	1,718	2026-04-22	apache-2.0	85
125	protectai/unbiased-toxic-roberta-onnx	protectai	token-classification	159,849	1,268,515	7	2026-07-09	apache-2.0	85
126	bigscience/bloom	bigscience	text-generation	0	4,862,625	5,033	2026-07-29	bigscience-bloom-rail-1.0	85
127	intfloat/multilingual-e5-small	intfloat	sentence-similarity	14,804,481	95,158,800	380	2026-04-02	mit	84
128	RedHatAI/gemma-4-31B-it-FP8-block	RedHatAI	image-text-to-text	4,500,032	9,174,208	44	2026-07-30	apache-2.0	84
129	ornith-ai/Ornith-1.0-35B-GGUF	ornith-ai	text-generation	3,384,148	4,326,113	1,017	2026-07-18	mit	84
130	Comfy-Org/MiniMax-H3	Comfy-Org	other	3,139,920	3,139,920	892	2026-08-06	other	84
131	Qwen/Qwen3.5-0.8B	Qwen	image-text-to-text	2,906,496	14,511,249	653	2026-03-02	apache-2.0	84
132	Qwen/Qwen3-Reranker-0.6B	Qwen	text-ranking	2,774,844	14,637,849	387	2026-04-16	apache-2.0	84
133	deepseek-ai/DeepSeek-V4-Flash	deepseek-ai	text-generation	2,577,975	9,504,049	2,055	2026-06-22	mit	84
134	google/gemma-4-31B-it-assistant	google	any-to-any	1,246,764	2,708,981	316	2026-07-15	apache-2.0	84
135	SamLowe/roberta-base-go_emotions	SamLowe	text-classification	904,023	97,748,993	685	2026-05-13	mit	84
136	Qwen/Qwen3.5-397B-A17B-FP8	Qwen	image-text-to-text	825,000	5,251,958	182	2026-04-24	apache-2.0	84
137	cyankiwi/Qwen3.5-4B-AWQ-4bit	cyankiwi	image-text-to-text	751,964	2,983,188	18	2026-07-21	apache-2.0	84
138	nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8	nvidia	text-generation	566,152	6,557,911	356	2026-03-15	other	84
139	lightonai/LightOnOCR-2-1B	lightonai	image-text-to-text	500,608	2,872,188	789	2026-07-08	apache-2.0	84
140	unsloth/Qwen3.6-35B-A3B-NVFP4-Fast	unsloth	image-text-to-text	447,417	447,417	97	2026-07-12	apache-2.0	84
141	cyankiwi/GLM-5.2-AWQ-INT4	cyankiwi	text-generation	395,948	433,482	14	2026-07-28	mit	84
142	zai-org/GLM-4.1V-9B-Thinking	zai-org	image-text-to-text	389,472	4,507,496	784	2026-07-22	mit	84
143	swiss-ai/Apertus-8B-Instruct-2509	swiss-ai	text-generation	374,478	2,933,999	484	2026-07-17	apache-2.0	84
144	litert-community/gemma-4-E4B-it-litert-lm	litert-community	other	371,261	1,550,498	187	2026-08-07	apache-2.0	84
145	bosonai/higgs-tts-3-4b	bosonai	text-to-speech	337,589	528,438	701	2026-07-09	other	84
146	protectai/xlm-roberta-base-language-detection-onnx	protectai	text-classification	156,819	1,027,697	6	2026-07-09	mit	84
147	cyankiwi/Qwen3-VL-4B-Instruct-AWQ-4bit	cyankiwi	image-text-to-text	154,336	1,287,521	9	2026-07-21	apache-2.0	84
148	Datadog/Toto-Open-Base-1.0	Datadog	time-series-forecasting	149,802	9,997,766	142	2026-05-14	apache-2.0	84
149	LiquidAI/LFM2-1.2B	LiquidAI	text-generation	0	3,000,497	363	2026-08-05	other	84
150	intfloat/multilingual-e5-base	intfloat	sentence-similarity	6,865,365	54,404,244	379	2026-04-02	mit	83
151	ibm-granite/granite-embedding-small-english-r2	ibm-granite	feature-extraction	3,740,585	15,042,484	75	2026-01-21	apache-2.0	83
152	Qwen/Qwen3-ASR-1.7B	Qwen	automatic-speech-recognition	2,568,541	10,528,073	992	2026-01-30	apache-2.0	83
153	ResembleAI/chatterbox	ResembleAI	text-to-speech	2,320,253	19,248,065	1,727	2026-06-10	mit	83
154	Lightricks/LTX-2.3	Lightricks	image-to-video	1,785,826	10,173,167	1,769	2026-08-02	other	83
155	Comfy-Org/Qwen-Image_ComfyUI	Comfy-Org	other	1,492,740	21,998,799	462	2026-06-06	apache-2.0	83
156	lmstudio-community/gemma-4-E4B-it-MLX-4bit	lmstudio-community	any-to-any	1,183,273	3,716,653	22	2026-07-23	apache-2.0	83
157	lmstudio-community/gemma-4-E4B-it-MLX-8bit	lmstudio-community	any-to-any	1,146,918	3,580,523	8	2026-07-23	apache-2.0	83
158	lmstudio-community/gemma-4-E4B-it-MLX-5bit	lmstudio-community	any-to-any	1,146,720	3,219,411	0	2026-07-23	apache-2.0	83
159	lmstudio-community/gemma-4-E4B-it-MLX-6bit	lmstudio-community	any-to-any	1,141,783	3,527,336	3	2026-07-23	apache-2.0	83
160	vcruz305/Hy3-GGUF	vcruz305	text-generation	879,149	891,172	18	2026-07-14	apache-2.0	83
161	microsoft/phi-4	microsoft	text-generation	663,989	13,379,801	2,286	2026-07-14	mit	83
162	lmstudio-community/gemma-4-12B-it-QAT-GGUF	lmstudio-community	other	509,130	1,336,074	13	2026-07-20	apache-2.0	83
163	Kijai/WanVideo_comfy_fp8_scaled	Kijai	other	481,658	7,797,717	725	2026-06-13	apache-2.0	83
164	microsoft/Mage-VL	microsoft	image-text-to-text	456,140	456,140	298	2026-08-06	apache-2.0	83
165	unsloth/gemma-4-E2B-it-GGUF	unsloth	image-text-to-text	434,893	3,664,171	280	2026-07-17	apache-2.0	83
166	nvidia/Qwen3.5-397B-A17B-NVFP4	nvidia	text-generation	338,508	2,799,404	105	2026-06-30	apache-2.0	83
167	lmstudio-community/gemma-4-E2B-it-MLX-4bit	lmstudio-community	any-to-any	205,523	622,868	1	2026-07-23	apache-2.0	83
168	lmstudio-community/gemma-4-E2B-it-MLX-8bit	lmstudio-community	any-to-any	197,619	589,786	1	2026-07-23	apache-2.0	83
169	lmstudio-community/gemma-4-E2B-it-MLX-6bit	lmstudio-community	any-to-any	196,527	582,407	0	2026-07-23	apache-2.0	83
170	lmstudio-community/gemma-4-E2B-it-MLX-5bit	lmstudio-community	any-to-any	196,524	498,180	0	2026-07-23	apache-2.0	83
171	google/gemma-4-E2B	google	any-to-any	0	1,931,282	421	2026-07-15	apache-2.0	83
172	sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2	sentence-transformers	sentence-similarity	57,173,935	536,808,143	1,340	2026-01-28	apache-2.0	82
173	openai/gpt-oss-20b	openai	text-generation	8,229,095	88,239,103	4,884	2025-08-26	apache-2.0	82
174	Qwen/Qwen3.5-4B	Qwen	image-text-to-text	6,547,546	34,614,170	796	2026-03-02	apache-2.0	82
175	ornith-ai/Ornith-1.0-9B-GGUF	ornith-ai	text-generation	4,567,553	5,443,384	614	2026-06-25	mit	82
176	google/gemma-4-12B-it-qat-w4a16-ct	google	any-to-any	1,586,352	4,096,559	50	2026-07-20	apache-2.0	82
177	google/gemma-4-31B	google	image-text-to-text	722,892	2,641,801	491	2026-07-15	apache-2.0	82
178	nvidia/llama-nemotron-rerank-1b-v2	nvidia	text-ranking	678,977	2,032,326	59	2026-05-20	other	82
179	HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive	HauhauCS	other	477,842	3,739,577	1,861	2026-06-05	apache-2.0	82
180	cyankiwi/Qwen3.6-27B-AWQ-BF16-INT4	cyankiwi	image-text-to-text	422,490	1,436,851	41	2026-07-21	apache-2.0	82
181	DeepBeepMeep/Wan2.1	DeepBeepMeep	other	384,728	3,885,780	46	2026-07-29	none	82
182	bosonai/higgs-tts-2-3b-base	bosonai	text-to-speech	377,419	3,836,510	694	2026-06-25	other	82
183	GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-Thinking-GGUF	GnLOLot	text-generation	347,087	349,065	323	2026-07-13	apache-2.0	82
184	nvidia/parakeet-tdt-0.6b-v3	nvidia	automatic-speech-recognition	280,075	1,786,371	1,033	2026-08-05	cc-by-4.0	82
185	baidu/Qianfan-OCR	baidu	image-text-to-text	269,309	1,264,144	1,195	2026-04-29	apache-2.0	82
186	unsloth/gemma-4-12B-it-qat-GGUF	unsloth	any-to-any	254,266	805,233	397	2026-07-17	apache-2.0	82
187	z-lab/Qwen3.6-35B-A3B-DFlash	z-lab	text-generation	223,304	556,324	286	2026-06-19	apache-2.0	82
188	mistralai/Ministral-3-14B-Instruct-2512	mistralai	other	210,293	1,814,495	312	2026-07-15	apache-2.0	82
189	lmstudio-community/gemma-4-26B-A4B-it-MLX-4bit	lmstudio-community	image-text-to-text	160,854	697,634	9	2026-07-23	apache-2.0	82
190	lmstudio-community/gemma-4-26B-A4B-it-MLX-6bit	lmstudio-community	image-text-to-text	153,743	552,717	1	2026-07-23	apache-2.0	82
191	MiniMaxAI/MiniMax-H3	MiniMaxAI	image-text-to-video	0	18,112	2,861	2026-08-06	other	82
192	Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-GGUF	Jackrong	image-text-to-text	0	1,372,388	347	2026-07-09	apache-2.0	82
193	ctheodoris/Geneformer	ctheodoris	fill-mask	0	4,087,157	305	2026-05-26	apache-2.0	82
194	farbodtavakkoli/OTel-LLM-E4B-IT	farbodtavakkoli	text-generation	3,992,619	7,486,954	0	2026-06-23	apache-2.0	81
195	ornith-ai/Ornith-1.0-35B	ornith-ai	text-generation	2,652,066	3,189,315	472	2026-06-25	mit	81
196	Qwen/Qwen3.5-27B	Qwen	image-text-to-text	2,563,438	15,912,374	1,028	2026-04-24	apache-2.0	81
197	Qwen/Qwen3-TTS-12Hz-1.7B-Base	Qwen	other	2,558,567	12,976,106	474	2026-01-23	apache-2.0	81
198	ornith-ai/Ornith-1.0-9B	ornith-ai	text-generation	2,264,074	2,566,699	509	2026-06-25	mit	81
199	docling-project/docling-layout-heron	docling-project	other	1,731,625	11,270,237	50	2026-02-09	apache-2.0	81
200	QuantTrio/Qwen3.6-35B-A3B-AWQ	QuantTrio	image-text-to-text	1,079,534	3,121,917	32	2026-04-17	apache-2.0	81
201	unsloth/Qwen3.5-9B-GGUF	unsloth	image-text-to-text	986,097	6,179,426	816	2026-03-02	apache-2.0	81
202	unsloth/Qwen3.6-35B-A3B-GGUF	unsloth	image-text-to-text	868,930	6,170,723	1,491	2026-04-20	apache-2.0	81
203	lmstudio-community/gemma-4-E4B-it-GGUF	lmstudio-community	other	624,527	4,229,848	60	2026-07-20	apache-2.0	81
204	handy-computer/Voxtral-Mini-4B-Realtime-2602-gguf	handy-computer	automatic-speech-recognition	410,798	533,544	1	2026-06-28	apache-2.0	81
205	coolthor/Huihui-Qwen3.6-35B-A3B-abliterated-FP8-DYNAMIC	coolthor	image-text-to-text	401,736	406,617	4	2026-07-30	apache-2.0	81
206	mlx-community/gpt-oss-20b-MXFP4-Q8	mlx-community	text-generation	338,504	7,109,499	84	2026-03-19	apache-2.0	81
207	LiquidAI/LFM2.5-1.2B-Instruct-GGUF	LiquidAI	text-generation	214,513	958,922	205	2026-08-05	other	81
208	lmstudio-community/gemma-4-26B-A4B-it-MLX-8bit	lmstudio-community	image-text-to-text	155,898	626,011	3	2026-07-23	apache-2.0	81
209	lmstudio-community/gemma-4-26B-A4B-it-MLX-5bit	lmstudio-community	image-text-to-text	152,914	440,444	0	2026-07-23	apache-2.0	81
210	XiaomiMiMo/MiMo-V2-Flash	XiaomiMiMo	text-generation	0	1,054,166	748	2026-07-09	mit	81
211	nyralabs/CrisperWhisper	nyralabs	automatic-speech-recognition	0	962,551	341	2026-07-22	cc-by-nc-4.0	81
212	cross-encoder/ms-marco-MiniLM-L4-v2	cross-encoder	text-ranking	10,365,349	56,637,623	27	2025-08-29	apache-2.0	80
213	MahmoudAshraf/mms-300m-1130-forced-aligner	MahmoudAshraf	automatic-speech-recognition	2,331,164	72,737,626	96	2026-04-15	cc-by-nc-4.0	80
214	Qwen/Qwen3.5-35B-A3B	Qwen	image-text-to-text	2,302,143	15,898,974	1,482	2026-04-24	apache-2.0	80
215	nvidia/Gemma-4-26B-A4B-NVFP4	nvidia	text-generation	1,455,036	5,309,305	126	2026-05-11	apache-2.0	80
216	RedHatAI/gemma-4-31B-it-NVFP4	RedHatAI	image-text-to-text	1,200,979	1,961,242	57	2026-07-30	apache-2.0	80
217	kingabzpro/wav2vec2-large-xls-r-300m-Urdu	kingabzpro	automatic-speech-recognition	861,112	13,582,639	14	2026-06-24	apache-2.0	80
218	cyankiwi/Qwen3-30B-A3B-Instruct-2507-AWQ-4bit	cyankiwi	text-generation	704,577	2,283,297	32	2026-07-21	apache-2.0	80
219	prism-ml/Ternary-Bonsai-27B-mlx-2bit	prism-ml	text-generation	662,839	662,845	168	2026-07-14	apache-2.0	80
220	unsloth/gemma-4-12b-it-GGUF	unsloth	image-text-to-text	576,071	2,227,064	788	2026-07-17	apache-2.0	80
221	Qwen/Qwen3.5-397B-A17B	Qwen	image-text-to-text	389,957	5,049,706	1,548	2026-04-24	apache-2.0	80
222	biohub/ESMFold2-Experimental-Fast	biohub	other	283,861	499,542	0	2026-07-28	mit	80
223	handy-computer/whisper-large-v3-turbo-gguf	handy-computer	automatic-speech-recognition	281,652	355,965	1	2026-07-21	apache-2.0	80
224	biohub/ESMFold2-Experimental-Fast-Cutoff2025	biohub	other	256,796	469,554	0	2026-07-28	mit	80
225	lmstudio-community/gemma-4-26B-A4B-it-QAT-GGUF	lmstudio-community	other	223,046	578,307	13	2026-07-20	apache-2.0	80
226	Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled	Jackrong	image-text-to-text	0	1,248,686	2,929	2026-07-07	apache-2.0	80
227	thinkingmachines/Inkling	thinkingmachines	image-text-to-text	0	76,073	1,697	2026-07-23	apache-2.0	80
228	PaddlePaddle/PaddleOCR-VL	PaddlePaddle	image-text-to-text	0	180,827	1,641	2026-08-05	apache-2.0	80
229	Alissonerdx/BFS-Best-Face-Swap	Alissonerdx	image-to-image	0	594,014	749	2026-07-30	mit	80
230	PaddlePaddle/PaddleOCR-VL-1.5	PaddlePaddle	image-text-to-text	0	565,872	659	2026-07-10	apache-2.0	80
231	Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-v2-GGUF	Jackrong	image-text-to-text	0	656,322	613	2026-07-09	apache-2.0	80
232	nvidia/Nemotron-Cascade-2-30B-A3B	nvidia	text-generation	0	513,806	522	2026-07-09	other	80
233	numind/NuMarkdown-8B-Thinking	numind	image-to-text	0	1,959,289	493	2026-06-05	mit	80
234	naver-hyperclovax/HyperCLOVAX-SEED-Think-32B	naver-hyperclovax	text-generation	0	616,075	403	2026-07-14	other	80
235	Jackrong/Qwopus3.6-27B-v2-MTP-GGUF	Jackrong	image-text-to-text	0	472,171	389	2026-07-09	apache-2.0	80
236	nvidia/canary-1b-flash	nvidia	automatic-speech-recognition	0	2,076,560	278	2026-06-29	cc-by-4.0	80
237	numind/NuExtract-1.5	numind	text-generation	0	1,414,781	247	2026-05-19	mit	80
238	Qwen/Qwen3-Embedding-0.6B	Qwen	feature-extraction	9,026,705	75,202,928	1,138	2026-04-20	apache-2.0	79
239	argmaxinc/whisperkit-coreml	argmaxinc	automatic-speech-recognition	8,174,912	57,693,231	199	2026-04-24	none	79
240	zai-org/GLM-5.2-FP8	zai-org	text-generation	2,368,604	4,689,414	248	2026-07-02	mit	79
241	Kijai/WanVideo_comfy	Kijai	other	1,730,689	69,406,609	2,469	2026-06-13	none	79
242	ornith-ai/Ornith-1.0-35B-FP8	ornith-ai	text-generation	879,726	1,068,455	81	2026-06-26	mit	79
243	openbmb/MiniCPM-V-4.6	openbmb	image-text-to-text	874,596	2,493,976	1,175	2026-07-01	apache-2.0	79
244	PaddlePaddle/PaddleOCR-VL-1.6-GGUF	PaddlePaddle	other	684,403	1,563,892	53	2026-06-10	apache-2.0	79
245	lmstudio-community/Qwen3.5-9B-MLX-8bit	lmstudio-community	image-text-to-text	567,918	1,406,464	1	2026-06-02	apache-2.0	79
246	fastino/gliner2-base-v1	fastino	other	494,268	2,557,435	96	2026-05-19	apache-2.0	79
247	google/gemma-4-E2B-it-qat-w4a16-ct	google	any-to-any	458,156	939,314	8	2026-07-20	apache-2.0	79
248	Abiray/Minimax-H3-nvfp4-INT4-INT8-Convrot	Abiray	image-text-to-video	452,420	452,420	121	2026-08-06	other	79
249	intfloat/e5-mistral-7b-instruct	intfloat	feature-extraction	419,911	7,282,903	569	2026-04-02	mit	79
250	ibm-granite/granite-speech-4.1-2b	ibm-granite	automatic-speech-recognition	402,320	1,585,627	158	2026-06-12	apache-2.0	79
251	DavidAU/Qwen3.5-9B-The-Defiant-Fable-Uncensored-Heretic-NEO-IMATRIX-MAX-MTP-GGUF	DavidAU	image-text-to-text	342,556	342,556	298	2026-08-07	apache-2.0	79
252	LuffyTheFox/Qwen3.6-35B-A3B-Uncensored-Genesis-Hermes-V7-GGUF	LuffyTheFox	image-text-to-text	332,992	332,992	416	2026-08-06	apache-2.0	79
253	lmstudio-community/gemma-4-31B-it-QAT-GGUF	lmstudio-community	other	201,674	526,327	7	2026-07-20	apache-2.0	79
254	mistralai/Ministral-3-3B-Instruct-2512-BF16	mistralai	other	175,812	589,240	34	2026-07-15	apache-2.0	79
255	lmstudio-community/gemma-4-12B-it-GGUF	lmstudio-community	other	172,193	788,016	25	2026-07-20	apache-2.0	79
256	LiquidAI/LFM2.5-8B-A1B	LiquidAI	text-generation	170,464	412,737	704	2026-08-04	other	79
257	ibm-granite/granite-vision-4.1-4b	ibm-granite	image-text-to-text	166,894	559,200	100	2026-07-13	apache-2.0	79
258	mistralai/Mistral-Small-4-119B-2603	mistralai	other	163,102	541,882	412	2026-07-15	apache-2.0	79
259	lj1995/VoiceConversionWebUI	lj1995	other	0	0	1,207	2026-08-01	mit	79
260	maya-research/maya1	maya-research	text-to-speech	0	340,551	891	2026-07-11	apache-2.0	79
261	nvidia/Alpamayo-R1-10B	nvidia	robotics	0	322,510	427	2026-08-04	openmdw-1.1	79
262	LiquidAI/LFM2.5-350M	LiquidAI	text-generation	0	401,792	393	2026-08-05	other	79
263	LiquidAI/LFM2-8B-A1B	LiquidAI	text-generation	0	392,262	371	2026-08-05	other	79
264	mistralai/Devstral-2-123B-Instruct-2512	mistralai	other	0	321,823	330	2026-07-15	other	79
265	poolside/Laguna-XS.2	poolside	text-generation	0	389,592	319	2026-07-14	apache-2.0	79
266	LoliRimuru/moeFussion	LoliRimuru	text-to-image	0	359,836	294	2026-07-14	creativeml-openrail-m	79
267	Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF	Jackrong	image-text-to-text	0	1,010,129	293	2026-07-04	apache-2.0	79
268	cross-encoder/ms-marco-MiniLM-L6-v2	cross-encoder	text-ranking	85,315,799	437,122,510	297	2025-08-29	apache-2.0	78
269	sentence-transformers/paraphrase-multilingual-mpnet-base-v2	sentence-transformers	sentence-similarity	11,111,585	118,348,637	486	2025-08-19	apache-2.0	78
270	Qwen/Qwen3-32B	Qwen	text-generation	7,645,702	50,072,602	728	2025-07-26	apache-2.0	78
271	Qwen/Qwen3.6-35B-A3B	Qwen	image-text-to-text	5,592,707	21,935,374	2,643	2026-04-24	apache-2.0	78
272	cyankiwi/gemma-4-26B-A4B-it-AWQ-4bit	cyankiwi	image-text-to-text	3,331,398	16,906,437	91	2026-07-21	apache-2.0	78
273	lmstudio-community/Qwen3.6-27B-MLX-8bit	lmstudio-community	image-text-to-text	917,599	2,361,265	14	2026-06-02	apache-2.0	78
274	lmstudio-community/Qwen3.6-27B-MLX-6bit	lmstudio-community	image-text-to-text	846,515	2,181,391	2	2026-06-02	apache-2.0	78
275	lmstudio-community/gemma-4-26B-A4B-it-QAT-MLX-4bit	lmstudio-community	image-text-to-text	844,724	2,309,592	12	2026-07-23	apache-2.0	78
276	lmstudio-community/Qwen3.6-27B-MLX-5bit	lmstudio-community	image-text-to-text	832,549	2,157,685	0	2026-06-02	apache-2.0	78
277	ornith-ai/Ornith-1.0-397B-FP8	ornith-ai	text-generation	615,913	781,801	180	2026-06-25	mit	78
278	nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16	nvidia	text-generation	493,863	672,636	305	2026-06-10	other	78
279	typhoon-ai/typhoon2.5-qwen3-4b	typhoon-ai	text-generation	457,450	701,695	6	2026-06-11	apache-2.0	78
280	speakleash/Bielik-11B-v3.0-Instruct	speakleash	text-generation	448,669	2,448,700	83	2026-07-01	apache-2.0	78
281	bartowski/Qwen_Qwen3.6-35B-A3B-GGUF	bartowski	image-text-to-text	427,072	896,002	139	2026-05-20	apache-2.0	78
282	Qwen/Qwen3.5-9B-Base	Qwen	image-text-to-text	408,424	1,129,344	97	2026-04-23	apache-2.0	78
283	cyankiwi/Devstral-Small-2-24B-Instruct-2512-AWQ-4bit	cyankiwi	other	388,037	1,067,610	14	2026-07-21	apache-2.0	78
284	SulphurAI/Sulphur-2-base	SulphurAI	text-to-video	386,064	2,941,369	1,968	2026-08-05	none	78
285	typhoon-ai/typhoon-ocr-3b	typhoon-ai	image-text-to-text	366,065	757,201	9	2026-06-11	apache-2.0	78
286	gravitee-io/bert-small-pii-detection	gravitee-io	token-classification	349,330	693,024	6	2026-05-21	apache-2.0	78
287	tabularisai/multilingual-sentiment-analysis	tabularisai	text-classification	329,648	5,913,853	393	2026-07-31	cc-by-nc-4.0	78
288	cyankiwi/Qwen3.5-9B-AWQ-4bit	cyankiwi	image-text-to-text	287,482	2,748,553	35	2026-07-21	apache-2.0	78
289	AtlasCloud/DeepSeek-V4-Flash-0731-FP8-DSpark	AtlasCloud	other	276,723	276,723	3	2026-07-31	mit	78
290	GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking-GGUF	GnLOLot	text-generation	269,077	269,077	182	2026-07-13	apache-2.0	78
291	unsloth/inkling-GGUF	unsloth	image-text-to-text	252,998	252,998	133	2026-07-16	apache-2.0	78
292	google/tipsv2-so400m14	google	zero-shot-image-classification	252,567	268,820	18	2026-07-28	apache-2.0	78
293	droplychee/droplychee-1.0-27b	droplychee	image-text-to-text	252,417	252,417	2	2026-08-05	apache-2.0	78
294	answerdotai/answerai-colbert-small-v1	answerdotai	other	250,948	55,222,361	160	2026-02-14	apache-2.0	78
295	thinkingmachines/Inkling-NVFP4	thinkingmachines	image-text-to-text	246,280	246,280	86	2026-07-30	apache-2.0	78
296	unsloth/Kimi-K3-GGUF	unsloth	image-text-to-text	245,703	245,703	326	2026-08-07	other	78
297	google/gemma-4-26B-A4B-it-assistant	google	any-to-any	244,012	815,071	174	2026-07-15	apache-2.0	78
298	OpenMOSS-Team/MOSS-Transcribe-Diarize	OpenMOSS-Team	audio-text-to-text	219,772	222,165	366	2026-07-31	apache-2.0	78
299	cyankiwi/gemma-4-26B-A4B-it-qat-AWQ-INT4	cyankiwi	image-text-to-text	216,971	296,122	8	2026-07-21	apache-2.0	78
300	unsloth/Ornith-1.0-35B-GGUF	unsloth	text-generation	204,529	204,529	133	2026-07-18	mit	78

How to use a Hugging Face model

Every model on this list can be loaded directly with the transformers library, using AutoModel.from_pretrained(“org/model-id”), or downloaded standalone with huggingface_hub’s snapshot_download. Click a model’s name to open its official Hugging Face page, where the model card documents exact usage, required libraries, and any license terms you need to accept first.

A short checklist before you commit to one in production:

1. Check the licence on the model card, not the tag. 7 models here state none at all, and “other” covers custom terms.
2. Prefer downloads over likes for infrastructure choices. The most-liked model on this list is #89 by downloads.
3. Check the last-modified date, remembering it moves on README edits too.
4. Match the format to your hardware. 41% of this list exists because the original weights did not fit somewhere; a GGUF or AWQ build may be what you actually want.
5. Watch for gated models if you are automating downloads, since they need terms accepted first.

What This Data Says About the Hugging Face Hub

Five conclusions, all checkable against the tables above:

1. The Hub runs on embeddings. 25 Embeddings & Retrieval models take 55.5% of downloads, roughly 13x the downloads per model of the most crowded category.
2. Downloads are a defaults game. One compact model from years ago holds 28.3% of all downloads, because it is what the tutorials use.
3. Attention and use are different axes. The most-liked model ranks #89 by downloads.
4. A large slice of the ecosystem is repackaging, not new models. 41% of the list is quantized or converted re-uploads.
5. Licensing is mostly settled and occasionally hazardous. 80% permissive, but 7 with no licence stated is the group that can actually cause you a problem.

“The number that reframed this for me is that one small embedding model from a few years ago pulls more downloads than every image generator, speech model and classifier on the list combined. Everyone argues about which chat model is best. Meanwhile the thing quietly running in production is a tiny sentence-transformer that got written into a tutorial once and never got replaced. If you want to know what an ecosystem actually depends on, count what it downloads, not what it upvotes.”
Alston Antony, founder of zplatform.ai and Senior Digital Marketing Manager at Brainstorm Force

Building AI into a product rather than picking a model? Our MCP servers directory ranks the connectors that let assistants act on real systems. For AI inside your site, see the best AI WordPress plugins, which carries a full CVE audit. For AI in the browser, the AI Chrome extensions and AI Firefox add-ons reports use the same data-first approach. More roundups sit in best AI tools.

Frequently Asked Questions

What are the best Hugging Face models?

By composite quality score the leader is amazon/chronos-2 at 98/100, and by raw downloads it is sentence-transformers/all-MiniLM-L6-v2 with 248,935,735 in 30 days. Which matters depends on the job: downloads indicate a proven, safe default, while the quality score weighs maintenance, growth and licensing alongside adoption. Both full rankings are above.

How is this different from Hugging Face’s own trending page?

Hugging Face’s trending view surfaces short-term spikes. This list weighs sustained adoption, how actively a model is maintained, and whether it is properly licensed, so it favours models genuinely in use over models briefly in the news. Hugging Face’s internal trending score is deliberately excluded from the ranking because its scale is not documented.

Are Hugging Face models free to use?

Most are. 80% of this list is Apache-2.0 or MIT, which permits commercial use with minimal conditions. But 7 models state no licence, which means no permission has been granted, and five are explicitly non-commercial. Always read the model card before shipping.

How many models are on the Hugging Face Hub?

Well over a million, the overwhelming majority of which are never downloaded by anyone but their author. This report ranks the 300 that clear an adoption floor of at least 1,000 downloads in 30 days or 50 likes, as of 7 August 2026.

What is the most downloaded Hugging Face model?

sentence-transformers/all-MiniLM-L6-v2, a sentence-similarity model, with 248,935,735 downloads in the last 30 days, which is 28.3% of all downloads across this list. It is a small embedding model rather than a chat model, because embedding models get pulled into every build of every search and RAG pipeline that uses them.

How often is this list updated?

The dataset is rebuilt from the Hugging Face Hub API and republished here, with the pull date shown at the top. Growth figures compare each model against a stored snapshot from roughly 30 days earlier, so they depend on that history being kept rather than on any single refresh.

Methodology and How to Cite This Data

Source: the free, public Hugging Face Hub API, which supplies downloads (30-day and all-time), likes, licence, author, gated status, task tag and last-modified date for every model, plus 5 retained snapshots used to compute the 30-day download trend.

Sample: 300 models, 878,230,936 downloads in 30 days, 7,461,254,367 all-time, 171,360 likes, 105 publishers, 8 task groups, 12 distinct licences.

Ranking formula: Quality Score = 40% adoption (downloads and likes, log-scaled) + 25% maintenance (how recently updated) + 20% growth (30-day downloads trend) + 15% trust (a stated licence, an identified author, and not being access-gated).

Inclusion: automatic, with an adoption floor of at least 1,000 downloads in the last 30 days or at least 50 likes. No hand-picking, and no model pays for placement.

Stated limitations: because of the adoption floor, quality scores sit in a narrow 78 to 98 band and should be read as an ordering within an already-filtered set. Growth is null for 66 models too new to have 30-day history, and those are scored neutrally rather than penalised. The Hub’s last-modified date moves on any repository change, including documentation, so it indicates attention rather than retraining. Hugging Face’s own trending score is reported by the API but excluded from the ranking because its scale is undocumented.

Cite as: zplatform.ai, “Best Hugging Face Models: The Complete Ranked List,” data pulled 7 August 2026.

Every number here traces to a row in the tables above, which is the point of publishing the formula next to the ranking. The best Hugging Face model for you depends on the task and the licence you can live with, and you should be able to check my working rather than take my word for it.