Organization
Hugging Face
Hugging Face is a developer platform company and inference provider headquartered in Brooklyn, New York32. This catalog lists 3 open model releases, 6 software projects, 1 dataset, and 1 evaluation tool from Hugging Face, under licenses including the Apache License 2.0 and the BigCode OpenRAIL-M v1 License Agreement.
Documented roles: Developer platform · Inference provider · Open-source steward
Hugging Face runs the Hugging Face Hub, which hosts Git-based repositories of models, datasets, and demo apps (Spaces), and sells hosted inference and enterprise services. It also maintains open-source machine-learning libraries such as Transformers.81011
Status: On September 3, 2026 NVIDIA announced that it has agreed to acquire Hugging Face. NVIDIA's Form 8-K states that it entered into a definitive agreement to acquire Hugging Face, Inc. on September 2, 2026, and that the transaction is expected to close in the first half of 2027, subject to customary closing conditions including required regulatory approvals. As of 2026-09-29 the acquisition is pending; this catalog has no source documenting that it has closed.67
Using this organization’s offerings
Intended users
Hugging Face offers a free Hub account for individuals, a PRO plan for individual users, and Team and Enterprise plans for organizations, along with paid compute such as Spaces hardware and Inference Endpoints.14
Hosted and self-hosted access
The Hub hosts Git-based model, dataset, and Spaces repositories that users can download from or upload to. Inference Providers routes hosted inference requests to partner providers at the providers' rates, with monthly credits that vary by plan. Hugging Face also maintains the Apache 2.0-licensed Transformers library.151718
Restrictions and review limits
Repository owners set licenses in each card's metadata, and the Hub documentation asks users to seek out and respect each project's license; a repository hosted on the Hub is not thereby a Hugging Face release. This profile is not a complete product or rights audit.1615
Key facts
- Legal name
- Hugging Face, Inc.1
- Ownership (separate from openness)
- Privately held
Recorded classification; fact-level ownership verification pending.
- Legal form
- Delaware corporation1
- Other documented locations
- Paris, France (Hugging Face SAS, the company's EU main establishment)2
- Founded
- 20164
- Sectors
- Open models, Data
What is and isn’t public
Hugging Face hosts public and private repositories for other developers' models and datasets and publishes open-source libraries, including Transformers under the Apache License 2.0. Hosted inference and enterprise plans are commercial services. This record does not assess any Hugging Face model release.811
Products
Hosted model products
Labeled Hosted product/API — this describes how a documented product is delivered (hosted access). It does not determine whether underlying weights or code are open; assess the specific release and its licenses.
| Product | Kind | Documented delivery and access | Reviewed |
|---|---|---|---|
| Inference Providers (external site: huggingface.co) | Hosted model API | Single API, including an OpenAI-compatible chat endpoint, that routes requests to partner inference providers and Hugging Face's own HF Inference; includes a free tier.9 | Sep 29, 2026 |
Other products
| Product | Kind | Documented delivery and access | Reviewed |
|---|---|---|---|
| Hugging Face Hub (external site: huggingface.co) | Developer tool | Web platform and API hosting public and private Git-based repositories for models, datasets, and Spaces apps; PRO, Team, and Enterprise subscription plans are documented.8 | Sep 29, 2026 |
| Inference Endpoints (external site: huggingface.co) | Inference service | Managed service for deploying a model from the Hub to production on fully managed, autoscaling infrastructure using engines such as vLLM, TGI, or a custom container.10 | Sep 29, 2026 |
Cataloged artifacts
Model families
- SmolLMModel family
- StarCoder2Model family
Model releases
- SmolLM3 3BModel releaseAvailability: PublicModel-disclosure tier (USASI rubric v0.2): Open-weight
- StarCoder2-15BModel releaseAvailability: PublicModel-disclosure tier (USASI rubric v0.2): Open-weight
- StarCoder2-3BModel releaseAvailability: PublicModel-disclosure tier (USASI rubric v0.2): Open-weight
Frameworks
- DiffusersSoftwareAvailability: Public
- Sentence TransformersSoftwareAvailability: Public
- smolagentsSoftwareAvailability: Public
- TransformersSoftwareAvailability: Public
- TRL (Transformers Reinforcement Learning)SoftwareAvailability: Public
Runtimes
- llama.cppSoftwareAvailability: Public
Datasets
- FineWebDatasetAvailability: Public
Evaluation tools
- BigCodeBenchEvaluation toolAvailability: Public
Notable documented facts
In the news
Amazon announces an Apache 2.0 release of TorchNeuron, a native PyTorch backend for AWS Trainium
Amazon said TorchNeuron, which it describes as running existing PyTorch code on its Trainium chips without a rewrite by plugging into PyTorch's standard mechanism for registering new hardware, would be released under the Apache 2.0 license at PyTorch Conference North America on October 20, 2026. Amazon says TorchNeuron supports Trainium2 and Trainium3 and that it is contributing changes upstream to PyTorch. The aws-neuron/torch-neuronx repository on GitHub is public, lists the Apache License 2.0, and describes itself as a native PyTorch backend for AWS Trainium. A Hugging Face blog post dated October 19 describes running Hugging Face models with it. USASI has not tested TorchNeuron.
TRL 1.15.0 released with a fused LM head enabled by default
On October 8, 2026 Hugging Face published version 1.15.0 of TRL on GitHub. The release notes say the SFT, DPO, KTO, GRPO, RLOO and Distillation trainers now score tokens with a fused LM head, a Triton kernel that avoids building the full logits tensor, and that it is on by default; scoring this way needs Triton on a GPU. The notes list breaking changes: a PEFT adapter on
lm_headnow raises an error, anduse_liger_kernel=Trueis deprecated in the GRPO, RLOO, DPO and KTO trainers and will be removed in v2.0.0. The release also adds selective activation checkpointing in SFT. USASI has not tested these changes.Diffusers 0.41.0 adds a Qwen-Image 2.1 pipeline and deprecates ONNX support
On October 6, 2026 Hugging Face published Diffusers 0.41.0 on GitHub. The release notes say it adds Qwen-Image 2.1, with text-to-image generation, image editing, native transparency, and LoRA training, and introduces tensor-parallel checkpoint loading in which each rank reads its own slice of sharded weights. The notes deprecate ONNX support in favor of Optimum and remove previously deprecated APIs, including the
LuminaText2ImgPipelineandLumina2Text2ImgPipelinealiases. They also say minor releases will now be coordinated around new model integrations, as in Transformers. Qwen-Image 2.1 is a third-party model and is not a catalog entry; the notes do not state its license.Transformers 5.19.0 adds the EmbeddingGemma2 multimodal embedding model
On October 6, 2026 Hugging Face published version 5.19.0 of its Transformers library on GitHub. The release notes add support for EmbeddingGemma2, which the notes describe as a multimodal embedding model from Google built on the Gemma 4 architecture that encodes text, images, audio, and video into a shared 768-dimensional vector space and uses Matryoshka Representation Learning so embeddings can be truncated to 512, 256, or 128 dimensions. The notes also list breaking changes: every mixture-of-experts model that computes router logits now returns them when
output_router_logits=True, the"paged|"prefix for SDPA and flash attention implementations is deprecated in favor of continuous batching on the regular functions, andOwlv2ForObjectDetection.embed_image_querynow selects the query box by objectness score. EmbeddingGemma2 itself is a Google model and is not a separate catalog entry; USASI has not tested these changes.llama.cpp server adds support for decision models through a /v1/systemone endpoint
On October 2, 2026 the ggml team announced on the Hugging Face blog that the llama.cpp server now supports decision models through a new /v1/systemone endpoint. Decision models score the options a caller supplies instead of generating text. The pull request that adds the endpoint (#29818) was merged into the llama.cpp repository the same day; it describes these models as wrappers around existing embedding-model architectures and adds a decision head and server handling for them. The announcement lists six supported models (Julia-1, Laya, Kev-4B, lev, OpenJev, and Clef), while the pull request itself covers five of them and lists Clef support as a planned follow-up.
What this catalog does not know
- Official careers page: not verified by this catalog.
Have a primary source? How to report a correction.
U.S. eligibility
Eligible · basis: U.S. headquarters
Dual-country assessment. Hugging Face has entities in both the United States and France. The contracting entity for its services is Hugging Face, Inc., a Delaware corporation, and its terms use New York law and New York courts. The privacy policy states that the company and its servers are located in the United States. It names Hugging Face SAS in Paris only as the company's "main establishment in the European Union" for GDPR purposes. Hugging Face's 2024 submission to NIST uses a Hugging Face, Inc. letterhead with a Brooklyn, New York address while describing the company as "based in the U.S. and France"; its 2025 submission to OSTP describes it as a U.S. company. AFP reporting from September 2026 describes its headquarters as American. No source documents the French entity as the parent or the principal place of business. On that basis the U.S. entity is treated as the headquarters. NVIDIA's pending acquisition (a U.S. company) is noted but not relied on.1234567
Sources
This listing is not an endorsement or a federal approval.
Share or cite this page
- X (opens X)
- Bluesky (opens Bluesky)
- LinkedIn (opens LinkedIn)
- Reddit (opens Reddit)
- Hacker News (opens Hacker News)
- Email (opens your email app)
- RSS updates
Cite this page
APA style
USASI. (2026, October 16). Hugging Face. United States of America Superintelligence. https://unitedstatesofamericasuperintelligence.com/companies/hugging-face/
BibTeX
@misc{usasi_hugging_face,
author = {{USASI}},
title = {Hugging Face},
year = {2026},
month = oct,
howpublished = {\url{https://unitedstatesofamericasuperintelligence.com/companies/hugging-face/}},
note = {United States of America Superintelligence. Last updated 2026-10-16}
}The date is when this page was last updated. Add the date you read it if your style needs one.
For maintainers: a “Listed in USASI” badge
Links to this page. It says the entry is listed, not endorsed or rated. About the badge
Markdown (for a README)
[](https://unitedstatesofamericasuperintelligence.com/companies/hugging-face/)
HTML
<a href="https://unitedstatesofamericasuperintelligence.com/companies/hugging-face/"><img src="https://unitedstatesofamericasuperintelligence.com/badges/listed.svg" alt="Listed in the USASI catalog" height="20"></a>