Hacker ApprovedDepartments

Hugging Face

The most recommended of 3 model hubs

People recommend it as the main place to find, download and run open AI models, with a large community. Main caveat: security incidents and throttled downloads.

People recommend it as the main place to find, download and run open AI models, with a large community. Main caveat: security incidents and throttled downloads.

154 people gave an opinion

recommend it
92
criticize it
62
approval share
60%
Each figure is 2 people. RecommendsCriticizes

Recommended in 12 of 13 months

  1. November 2025: recommended
  2. December 2025: recommended
  3. January 2026: recommended
  4. February 2026: recommended
  5. March 2026: recommended
  6. April 2026: recommended
  7. May 2026: recommended
  8. June 2026: recommended
  9. July 2026: recommended
  10. August 2026: recommended
  11. September 2026: recommended
  12. October 2026: recommended
  13. November 2026: not recommended (month in progress)

Why it doesn’t carry the badge

  • 92 people recommend itNeeds 10 or more, met.
  • 1.4 recommenders for every criticNeeds 2 to 1 or better, not met.
  • Recommended in 12 of 13 monthsNeeds 4 or more, met.

The same three rules apply to every product, and the badge can’t be bought. How the badge works

What people say about it

The reasons people give, with the number of comments that give each one. Open a reason to read what they said.

Why people recommend it

Each mark is one comment

  • Main place to find and download models

    “I am not aware of any alternative for choosing and downloading the right model for local inference that is superior to using directly the Huggingface site”

    Comment from 16 Apr 2026 in “Stop Using Ollama”

    “I went with Hugging Face mainly because it is simple and already handles distribution and versioning”

    Comment from 19 Mar 2026 in “Show HN: Hacker News archive (47M+ items, 11.6GB) as Parquet, updated every 5m”

    “One good way to find inference providers for open models is through hugging face”

    Comment from 31 Jul 2026 in “Ask HN: What are you using for LLM inference in production?”
  • Trusted open model community

    “HF has become the front runner in open model hosting. They have a lot of trust from the wider community.”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

    “nothing else has been as effective at accelerating model development and research sharing”

    Comment from 11 Sep 2026 in “HuggingFace: Security.txt”

    “building such a huge AI community around a single place is probably impossible by now”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”
  • Free or cheap model hosting

    “Freemium seems to be working pretty well for them—what's the alternative website, after all. They seem to command their niche.”

    Comment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

    “there's places (hugging face?) that will host and run your updated model for cheaper than you could run it yourself”

    Comment from 22 Dec 2025 on Hacker News

    “who already did all the setup for you and just waits for your credit card per minute/seconds/token”

    Comment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”
  • Useful libraries and tooling

    “they also have training, inference hosting, and they develop/maintain a bunch of core AI infrastructure software”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

    “HF is super valuable with their transformers library, their code, tutorials, smol-models, etc”

    Comment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

    “Getting stuff to run before huggingface and uv and docker containers with cuda was way worse.”

    Comment from 10 Mar 2026 in “Open Weights Isn't Open Training”

What people warn about

Each mark is one comment

  • Security incidents and hacks

    “It is extremely embarrassing to get owned because you are letting arbitrary programs/users call out to the open web from the infra”

    Comment from 15 Sep 2026 in “Dario, Please”

    “If you think otherwise, look no further than the Huggingface hack and what happens when they run completely autonomously.”

    Comment from 16 Oct 2026 in “Ask HN: Thoughts on hiring junior engineers in 2026?”

    “HuggingFace wasn't doing at all while also allowing any user's arbitrary programs to call out to the open web from prod”

    Comment from 18 Sep 2026 in “A heap overflow and SSO misconfiguration to compromise OpenAI internal repos”
  • Lock-in and control after acquisitions

    “This acquisition will likely solidify this general stance, reduce free compute allowance for a subscription, cap downloads”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

    “now the company that sells the GPUs also owns the registry that decides which models are easy to find and download”

    Comment from 28 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

    “Combined with recent acquisition of OpenRouter we see exactly what we were increasingly concerned about”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”
  • Gated downloads, throttling and telemetry

    “an onerous license agreement that must be agreed to during registration and a bloated cli interface with loads of telemetry”

    Comment from 4 Sep 2026 in “Georgi Gerganov on llama.cpp/ggml future after Nvidia acquisition of HuggingFace”

    “Previously with non-gated models, you could download the XET files with any program, no special software needed.”

    Comment from 16 Oct 2026 in “Muse Spark Open”

    “pulling models is an absolute nightmare when I basically have to use HF and even with a token getting throttled”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”
  • Inference hosting feels unserious

    “Their inference hosting is a joke (I don't think I ever had a working response from a "try this model" form).”

    Comment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

    “Huggingface for a serious service has never felt truly serious to me for reasons like these.”

    Comment from 1 Aug 2026 in “Tailscale didn't stop the Hugging Face intrusion”

Other model hubs people recommend

The two most recommended other model hubs, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.

Hugging Face compared with 2 alternatives
ProductRecommendCriticizeApprovalMonthsOwnedBadge
Hugging FaceThis page926260%12——
ModelScope90100%5——
Civitai7558%5——
See all 3 model hubs, ranked

What people said

40 of the 163 comments that mention it, newest first. Each links to the original comment.

“If you think otherwise, look no further than the Huggingface hack and what happens when they run completely autonomously.”

MixedComment from 16 Oct 2026 in “Ask HN: Thoughts on hiring junior engineers in 2026?”

“Previously with non-gated models, you could download the XET files with any program, no special software needed.”

CriticizesComment from 16 Oct 2026 in “Muse Spark Open”

“if not for arxiv and huggingface, the field of ML would be nowhere near where it is today”

RecommendsComment from 24 Sep 2026 in “ArXiv receives multiyear commitments to support it as an independent nonprofit”

“How they host stuff and then leverage that position to build add-ons and lock-in, lobby for laws.”

CriticizesComment from 21 Sep 2026 in “Nobody pays for FOSS, we can force them to”

“HuggingFace wasn't doing at all while also allowing any user's arbitrary programs to call out to the open web from prod”

CriticizesComment from 18 Sep 2026 in “A heap overflow and SSO misconfiguration to compromise OpenAI internal repos”

“It is extremely embarrassing to get owned because you are letting arbitrary programs/users call out to the open web from the infra”

CriticizesComment from 15 Sep 2026 in “Dario, Please”

“nothing else has been as effective at accelerating model development and research sharing”

RecommendsComment from 11 Sep 2026 in “HuggingFace: Security.txt”

“an onerous license agreement that must be agreed to during registration and a bloated cli interface with loads of telemetry”

CriticizesComment from 4 Sep 2026 in “Georgi Gerganov on llama.cpp/ggml future after Nvidia acquisition of HuggingFace”

“use 3rd party marketplaces. cerebras is resold on vercel, openrouter and huggingface.”

RecommendsComment from 4 Sep 2026 in “Qwen 3.8 27B available on Cerebras at 1500 tokens/s”

“now the company that sells the GPUs also owns the registry that decides which models are easy to find and download”

CriticizesComment from 28 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“providing a sufficiently useful and general service that makes technology more accessible to all”

RecommendsComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“I hooked up the settings to download models from Hugging Face conveniently”

RecommendsComment from 27 Aug 2026 in “Small Models Have Arrived”

“HF has become the front runner in open model hosting. They have a lot of trust from the wider community.”

RecommendsComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“Their inference hosting is a joke (I don't think I ever had a working response from a "try this model" form).”

CriticizesComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“HF has been a close part of my ML/AI career, coinciding exactly when I moved into this space 10 years ago.”

RecommendsComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“building such a huge AI community around a single place is probably impossible by now”

RecommendsComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“Combined with recent acquisition of OpenRouter we see exactly what we were increasingly concerned about”

CriticizesComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“This acquisition will likely solidify this general stance, reduce free compute allowance for a subscription, cap downloads”

CriticizesComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“pulling models is an absolute nightmare when I basically have to use HF and even with a token getting throttled”

CriticizesComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“Brilliant replacement of http://build.nvidia.com to increase developer reach and deployment of nvidia open models”

RecommendsComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“they also have training, inference hosting, and they develop/maintain a bunch of core AI infrastructure software”

RecommendsComment from 27 Aug 2026 in “Nvidia agrees to acquire Hugging Face for $13B”

“the pricing of tokens on places like HuggingFace and OpenRouter suggests that probably isn't true”

MixedComment from 3 Aug 2026 in “Qwen3.8-Max: A New Bar for Coding and Cowork”

“Huggingface for a serious service has never felt truly serious to me for reasons like these.”

CriticizesComment from 1 Aug 2026 in “Tailscale didn't stop the Hugging Face intrusion”

“One good way to find inference providers for open models is through hugging face”

RecommendsComment from 31 Jul 2026 in “Ask HN: What are you using for LLM inference in production?”

“I am not aware of any alternative for choosing and downloading the right model for local inference that is superior to using directly the Huggingface site”

RecommendsComment from 16 Apr 2026 in “Stop Using Ollama”

“who already did all the setup for you and just waits for your credit card per minute/seconds/token”

RecommendsComment from 16 Apr 2026 in “Darkbloom – Private inference on idle Macs”

“Huggingface, Swartz et al have done more social/political good for this world than billions have.”

RecommendsComment from 12 Apr 2026 in “AI Will Be Met with Violence, and Nothing Good Will Come of It”

“The red/yellow/green indicators on huggingface.co are kind of nice, but you only know for sure when you try to load the model”

MixedComment from 2 Apr 2026 in “Google releases Gemma 4 open models”

“I went with Hugging Face mainly because it is simple and already handles distribution and versioning”

RecommendsComment from 19 Mar 2026 in “Show HN: Hacker News archive (47M+ items, 11.6GB) as Parquet, updated every 5m”

“Getting stuff to run before huggingface and uv and docker containers with cuda was way worse.”

RecommendsComment from 10 Mar 2026 in “Open Weights Isn't Open Training”

“The other nice part of huggingface's setup is you can add theoretical hardware and search that way too.”

RecommendsComment from 2 Mar 2026 in “Right-sizes LLM models to your system's RAM, CPU, and GPU”

“And I said all of that despite us continuing to use their platform and libraries extensively…”

MixedComment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

“Freemium seems to be working pretty well for them—what's the alternative website, after all. They seem to command their niche.”

RecommendsComment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

“HF is super valuable with their transformers library, their code, tutorials, smol-models, etc”

RecommendsComment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

“one of the few quiet heroes (along with Chinese OSS) helping bring on-premise AI to the masses”

RecommendsComment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

“Hugging Face has large number of coding models that you can download for free”

RecommendsComment from 9 Feb 2026 in “No one on Hacker News can answer this question”

“benchmark the models you could run on your potential hardware on (neo) cloud provider like HuggingFace”

RecommendsComment from 15 Jan 2026 in “Apple Is Fighting for TSMC Capacity as Nvidia Takes Center Stage”

“there's places (hugging face?) that will host and run your updated model for cheaper than you could run it yourself”

RecommendsComment from 22 Dec 2025 on Hacker News

“There are plenty of models on HuggingFace that I can run on my Mac Studio”

RecommendsComment from 28 Nov 2025 in “A trillion dollars (potentially) wasted on gen-AI”

“All the data used by Apertus is just data processed or generated by American companies(NVidia, Apple and huggingface mostly).”

MixedComment from 21 Nov 2025 in “Olmo 3: Charting a path through the model flow to lead open-source AI”

Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.