Products at a glance
huggingface.co names 18 products under Hugging Face.
- LeRobot — Python library providing state-of-the-art machine learning tooling and datasets for real-world robotics.
- smolagents — Open-source Python library for building and running agents in just a few lines of code.
- Transformers — Model-definition framework for state-of-the-art text, vision, audio, video and multimodal ML models.
- Reachy Mini — Open-source expressive desktop robot for hackers and AI builders, with wireless and Lite versions.
- huggingface_hub — Python client library for interacting with the Hugging Face Hub from code.
- Lighteval — All-in-one toolkit for evaluating large language models across multiple inference backends.
- Optimum — Extension of Transformers providing performance optimization tools for targeted hardware.
- AutoTrain — No-code platform by Hugging Face for training state-of-the-art NLP, computer vision and tabular models.
- OpenEnv — Unified framework for building and deploying isolated execution environments for agentic reinforcement learning.
- Xet — Content-addressed storage protocol specification for chunking, deduplicating and reconstructing files on the Hub.
- Optimum TPU — Archived Hugging Face package for optimizing AI workloads on Google Cloud TPU devices.
- Inference Endpoints by Hugging Face — Deploy popular AI models in minutes using Inference Endpoint.
- PRO Account — Hugging Face personal paid subscription priced at $9 per month.
- Team — Hugging Face organisation plan at $20 per user per month with enterprise-grade controls.
- Spaces — Interactive apps on the Hub for demonstrating ML models directly in the browser.
- Enterprise — Hugging Face enterprise plan at $50 per user per month, sold via sales team.
- Evaluate — Hugging Face documentation and tooling for evaluating AI models on the Hub via leaderboards, model cards and libraries.
- Storage Buckets — Non-versioned, mutable object storage offered on the Hugging Face Hub alongside Git-based repositories.
LeRobot
LeRobot provides a hardware-agnostic, Python-native interface for controlling real robots, from affordable arms like the SO-ARM101 to full humanoids, plus tools to record, store and share datasets. Every dataset uses the standardized LeRobotDataset format and can be streamed from the Hugging Face Hub. It implements state-of-the-art policies from imitation-learning models like ACT to vision-language-action models like π₀ and SmolVLA, following a Teleoperate → Record → Train → Deploy workflow.
smolagents
smolagents is an open-source Python library designed to make it extremely easy to build and run agents with a few lines of code, with agent logic fitting in roughly a thousand lines. It offers first-class support for Code Agents (CodeAgent) with sandboxed execution via Modal, Blaxel, E2B or Docker, plus ToolCallingAgent for JSON/text tool-calling. It is model-agnostic, modality-agnostic and tool-agnostic, with Hub integrations to share agents and tools as Gradio Spaces.
Transformers
Transformers acts as the model-definition framework for state-of-the-art machine learning models across text, computer vision, audio, video and multimodal tasks, for both inference and training. It is the pivot across frameworks, compatible with training frameworks such as Axolotl, Unsloth, DeepSpeed and FSDP, and inference engines like vLLM, SGLang and TGI. Main features include Pipeline, Trainer and generate. The page “Transformers · Hugging Face” on huggingface.co records: Transformers documentation Transformers Transformers Search documentation Join the Hugging Face community and get access to the augmented documentation experience Collaborate on models, datasets and Spaces Faster examples with accelerated inference Switch between documentation themes to get started Copy page Transformers Transformers acts as the model-definition framework for state-of-the-art….
Reachy Mini
Reachy Mini is an open-source, expressive robot made for hackers and AI builders, available as Reachy Mini (Wireless) with Raspberry Pi CM4, battery and WiFi, or Reachy Mini Lite which runs on your PC via USB. A simulation option lets users prototype in MuJoCo with no hardware. It comes with an app store powered by Hugging Face Spaces, including a Conversation App, Radio and Hand Tracker, and offers Python and JavaScript SDKs.
huggingface_hub
The huggingface_hub library allows you to interact with the Hugging Face Hub, a machine learning platform for creators and collaborators. It provides a simple way in Python to discover pre-trained models and datasets, create and share your own, manage repositories, take part in discussions and run inference. Documentation includes a quick start guide, how-to guides, reference and conceptual guides. The page “🤗 Hub client library · Hugging Face” on huggingface.co records: Hub Python Library documentation 🤗 Hub client library Hub Python Library Search documentation Join the Hugging Face community and get access to the augmented documentation experience Collaborate on models, datasets and Spaces Faster examples with accelerated inference Switch between documentation themes to get started Copy page 🤗 Hub client library The huggingface_hub library allows you to….
Lighteval
Lighteval is an all-in-one toolkit for evaluating Large Language Models across multiple backends, saving detailed sample-by-sample results for debugging and comparison. Supported backends include inspect-ai, transformers with Accelerate, Nanotron, VLLM, SGLang, Inference Endpoints, TGI, LiteLLM and Inference Providers. It ships thousands of pre-built evaluation tasks and supports custom tasks and metrics. The page “Lighteval · Hugging Face” on huggingface.co records: Lighteval documentation Lighteval Lighteval Search documentation Join the Hugging Face community and get access to the augmented documentation experience Collaborate on models, datasets and Spaces Faster examples with accelerated inference Switch between documentation themes to get started Copy page Lighteval 🤗 Lighteval is your all-in-one toolkit for evaluating Large Language Models (LLMs)….
Optimum
Optimum is an extension of Transformers that provides performance optimization tools to train and run models on targeted hardware with maximum efficiency. It is distributed as a collection of packages for hardware partners including NVIDIA, AWS Trainium/Inferentia, Google TPUs, Intel, Intel Gaudi and FuriosaAI. It also supports open-source frameworks such as ONNX Runtime and ExecuTorch. The page “🤗 Optimum · Hugging Face” on huggingface.co records: Optimum documentation 🤗 Optimum Optimum Search documentation You are viewing main version, which requires installation from source . If you'd like regular pip install, checkout the latest stable version ( v1.27.0 ).
AutoTrain
AutoTrain Advanced, developed by Hugging Face, is a no-code platform to simplify training state-of-the-art models across NLP, Computer Vision and Tabular Data. It can be used by creating a new space or through a more hands-on approach. The documentation notes the project is no longer maintained and recommends Axolotl, TRL or transformers.Trainer instead. The page “AutoTrain · Hugging Face” on huggingface.co records: AutoTrain documentation AutoTrain AutoTrain Search documentation You are viewing main version, which requires installation from source . If you'd like regular pip install, checkout the latest stable version ( v0.8.24 ).
OpenEnv
OpenEnv is a unified framework for building, deploying and interacting with isolated execution environments for agentic reinforcement learning, powered by Gymnasium-style APIs. It features container-first design, HTTP-native deployment, secure sandboxed isolation, a rich pre-built environment library and CLI tools. Getting started materials include a 5-part series and a Build Your Own Environment reference guide. The page “OpenEnv: Agentic Execution Environments · Hugging Face” on huggingface.co records: OpenEnv documentation OpenEnv: Agentic Execution Environments OpenEnv Search documentation You are viewing main version, which requires installation from source . If you'd like regular pip install, checkout the latest stable version ( v0.10.0 ).
Xet
The Xet Protocol Specification (version 1.1.0) defines the end-to-end protocol for content-addressed data, covering chunking and hashing rules, deduplication strategy, xorb and shard object formats, file reconstruction, authentication and the CAS APIs for upload and download. Its goal is interoperability and determinism so independent implementations produce the same hashes and objects. Reference implementations include xet-core (hf-xet + git-xet) and huggingface.js.
Optimum TPU
Optimum TPU provides the machinery to leverage and optimize AI workloads running on Google Cloud TPU devices, optimizing transformers and TGI and integrating the Hugging Face Hub. It offers the same user experience as Transformers with minimal changes required, and is pip installable. The documentation notes that Optimum-TPU is now archived and points users to TPU inference and Accelerate. The page “🤗 Optimum TPU · Hugging Face” on huggingface.co records: Google TPUs documentation 🤗 Optimum TPU Google TPUs Search documentation Join the Hugging Face community and get access to the augmented documentation experience Collaborate on models, datasets and Spaces Faster examples with accelerated inference Switch between documentation themes to get started Copy page 🤗 Optimum TPU **🚧 Optimum-TPU is now archived.** While this project is no longer under….
Sources
-
Welcome to the Hugging Face Audio course! · Hugging Face https://huggingface.co/learn/audio-course
-
Welcome to the 🤗 AI Agents Course · Hugging Face https://huggingface.co/learn/agents-course
-
🤗 Optimum · Hugging Face https://huggingface.co/docs/optimum
-
AutoTrain · Hugging Face https://huggingface.co/docs/autotrain
-
Reachy Mini 🤖 · Hugging Face https://huggingface.co/docs/reachy_mini
-
Lighteval · Hugging Face https://huggingface.co/docs/lighteval
-
Welcome to the Community Computer Vision Course · Hugging Face https://huggingface.co/learn/computer-vision-course
-
Hugging Face Hub documentation · Hugging Face https://huggingface.co/docs/hub
-
Evaluate on the Hub · Hugging Face https://huggingface.co/docs/evaluate
-
LeRobot · Hugging Face https://huggingface.co/docs/lerobot
-
Inference Endpoints · Hugging Face https://huggingface.co/docs/inference-endpoints
-
🤗 Optimum TPU · Hugging Face https://huggingface.co/docs/optimum-tpu
-
OpenEnv: Agentic Execution Environments · Hugging Face https://huggingface.co/docs/openenv
-
🤗 Hub client library · Hugging Face https://huggingface.co/docs/huggingface_hub
-
smolagents · Hugging Face https://huggingface.co/docs/smolagents
-
Transformers · Hugging Face https://huggingface.co/docs/transformers
-
Welcome to the 🤗 Robotics Course · Hugging Face https://huggingface.co/learn/robotics-course
Last verified 15 Nov 2026. This entry is compiled from the public web pages listed above. Nothing here is stated that those pages do not, and each of them was read on the date shown.