Hugging Face

Unsecured OpenAI agents posted 53 user images on the internetSep 26, 2026 · SecurityIncident → Multi-Region training with Amazon SageMaker HyperPod and QumuloSep 24, 2026 · PolicyChange → Accelerating vision-language models with LFM2.5-VL-DSparkSep 24, 2026 · HardwareLaunch → How to Use NVIDIA Warp and MjWarp to Accelerate Robotics Simulation and Learning WorkflowsSep 23, 2026 · Funding → Accelerate multimodal RL training with SkyRL on Amazon SageMaker HyperPodSep 22, 2026 · OpenSourceRelease → We Should Be Able to Change Our LanguagesSep 22, 2026 · ProductLaunch → Transformers now runs llama.cpp quantsSep 22, 2026 · OpenSourceRelease → Jun Kim, oMLX creator and maintainer, joins Hugging Face to support the MLX communitySep 22, 2026 · ProductUpdate → Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization ProblemSep 21, 2026 · Other → tokenizers v1: encode, decode and scaling, measuredSep 21, 2026 · ProductUpdate → Python Workers are now generally availableSep 18, 2026 · ProductUpdate → Your Agent Aced the Task. Will It Do It Again?Sep 15, 2026 · Research → Async GRPO with LoRA across HF Jobs: a bucket, a proxy, and no NCCLSep 10, 2026 · PolicyChange → Rebuilding AUTOMATIC1111 with Gradio WorkflowSep 10, 2026 · ProductLaunch → NeoMME: an efficient Multimodal-native and Multilingual EncoderSep 3, 2026 · Research → Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO StepsSep 3, 2026 · PolicyChange → Training a coding model to paint watercolours with TRL and OpenEnvSep 3, 2026 · Research → BenchMIRT: What are LLM benchmarks actually measuring?Sep 1, 2026 · Research → Project HydraFusion: Frontier quality via multi-model orchestrationSep 1, 2026 · PriceChange → Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AISep 1, 2026 · HardwareLaunch → The Open ASR Leaderboard Adds Its First Global South LanguageAug 28, 2026 · ProductLaunch → Training and Finetuning Multi-Vector Embedding Models with Sentence TransformersAug 26, 2026 · OpenSourceRelease → Granite 4.2 LLMs: How They're BuiltAug 25, 2026 · Acquisition → Quantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision originalAug 25, 2026 · ProductUpdate → Wire It, Run It, Deploy It: AI Workflows in GradioAug 25, 2026 · ProductLaunch → How Hugging Face Inference Endpoints, Jobs, and Buckets Power Search on Papers with CodeAug 21, 2026 · Research → Measuring benchmark optimization in speech recognitionAug 21, 2026 · Research → Up to 3.2x Faster Inference with LFM2.5-DSparkAug 20, 2026 · HardwareLaunch → How Much Memory Does Your Agent Actually Need?Aug 18, 2026 · OpenSourceRelease → Multi-Vector (Late Interaction) Embedding Models with Sentence TransformersAug 18, 2026 · ProductUpdate → Same Cluster, 33 Points More Utilization: What Changed Was the OrderAug 17, 2026 · PolicyChange → Record, train, and deploy from one place with Strands Agents, LeRobot, and Hugging Face Storage BucketsAug 13, 2026 · OpenSourceRelease → Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysisAug 12, 2026 · Acquisition → Thinking of ACE? We Can Do It with Fewer TokensAug 11, 2026 · ProductUpdate → Build Low-Latency Multilingual Voice Agents: Open Weights & Full Deployment Control with NVIDIA Magpie TTSAug 10, 2026 · ProductLaunch → v0.32.8Aug 10, 2026 · ProductLaunch → Making Knowledge Distillation Cheap Enough to Run at ScaleAug 10, 2026 · OpenSourceRelease → Meta is back with Muse Glimmer: local, agentic, multimodal, and open sourceAug 10, 2026 · Research → Baseten on Hugging Face Inference Providers 🔥Aug 6, 2026 · ProductLaunch → GPU Management: Why Idle GPUs Are the New Grounded AircraftJul 30, 2026 · PriceChange → NVIDIA Cosmos-H-Dreams: Bringing Real-Time Generative Simulation to Surgical RoboticsJul 27, 2026 · Research → Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 IncidentJul 27, 2026 · SecurityIncident → Bringing Nunchaku 4-bit Diffusion Inference to DiffusersJul 23, 2026 · ModelRelease → Grabette: an open system to record robot-manipulation dataJul 21, 2026 · PolicyChange → Newer Models, Same AdvantageJul 16, 2026 · SecurityIncident → Security incident disclosure — July 2026Jul 16, 2026 · SecurityIncident → Model Routing Is Simple. Until It Isn’t.Jul 15, 2026 · PriceChange → Introducing Real World VoiceEQ: Measuring the human quality of voice AIJul 15, 2026 · ProductUpdate → Welcome Inkling by Thinking MachinesJul 15, 2026 · PolicyChange → Profiling in PyTorch (Part 3): Attention is all you profileJul 10, 2026 · ProductLaunch → From Hugging Face to Amazon SageMaker Studio in one clickJul 7, 2026 · LeadershipChange → Native-speed vLLM transformers modeling backendJul 7, 2026 · ModelRelease → LeRobot v0.6.0: Imagine, Evaluate, ImproveJul 7, 2026 · Partnership → Run AI workloads on any cloud, store on Hugging Face: zero-egress storage with SkyPilotJul 7, 2026 · Partnership → PRX Part 4: Our Data StrategyJul 6, 2026 · Other → 🤗 Kernels: Major UpdatesJul 6, 2026 · HardwareLaunch → Hugging Face and Cerebras bring Gemma 4 to real-time voice AIJul 1, 2026 · Partnership → ScarfBench: Benchmarking AI Agents for Enterprise Java Framework MigrationJun 30, 2026 · Research → Why Specialization Is InevitableJun 30, 2026 · Acquisition → Featuring Every Eval Ever Results on Hugging Face Model PagesJun 30, 2026 · Research → DiScoFormer: One transformer for density and score, across distributionsJun 29, 2026 · Research → Run a vLLM Server on HF Jobs in One CommandJun 26, 2026 · ProductLaunch → Accelerating Transformers Fine-Tuning with NVIDIA NeMo AutoModelJun 24, 2026 · ProductUpdate → Introducing the FFASR Leaderboard: Benchmarking ASR in the Real WorldJun 24, 2026 · ProductLaunch → Shipping huggingface_hub every week with AI, open tools, and a human in the loopJun 23, 2026 · ProductUpdate → Experimenting with the proposed Cross-Origin Storage API in Transformers.jsJun 23, 2026 · OpenSourceRelease → PP-OCRv6 on Hugging Face: 50-Language OCR from 1.5M to 34.5M ParametersJun 22, 2026 · ProductLaunch → We got local models to triage the OpenClaw repo for FREE!*Jun 22, 2026 · Partnership → MosaicLeaks: Can your research agent keep a secret?Jun 18, 2026 · Research → Is it agentic enough? Benchmarking open models on your own toolingJun 18, 2026 · ProductLaunch → Beyond LoRA: Can you beat the most popular fine-tuning technique?Jun 18, 2026 · Partnership → Agentic Resource Discovery: Let agents searchJun 17, 2026 · ProductUpdate → v0.30.9Jun 15, 2026 · ProductLaunch → v0.23.0Jun 15, 2026 · Funding → Profiling in PyTorch (Part 2): From nn.Linear to a Fused MLPJun 11, 2026 · ProductLaunch → Migrating Your GitHub CI to Hugging Face JobsJun 9, 2026 · HardwareLaunch → How an Agent Built a 3D Paris Gallery by Chaining Two Hugging Face SpacesJun 8, 2026 · Other → Nemotron 3.5 Content Safety: Customizable Multimodal Safety for Global Enterprise AIJun 4, 2026 · PolicyChange → Direct Preference Optimization Beyond ChatbotsJun 3, 2026 · Research → Holo3.1: Fast & Local Computer Use AgentsJun 2, 2026 · Partnership → Introducing Mellum2: A 12B Mixture-of-Experts Model by JetBrainsJun 1, 2026 · Research → Designing the hf CLI as an agent-optimized way to work with the HubJun 1, 2026 · Research → Profiling in PyTorch (Part 1): A Beginner's Guide to torch.profilerMay 29, 2026 · ProductLaunch → Reachy Mini goes fully localMay 25, 2026 · Funding → OlmoEarth v1.1: A more efficient family of Earth observation modelsMay 19, 2026 · Research → Introducing the Ettin Reranker FamilyMay 19, 2026 · OpenSourceRelease → PaddleOCR 3.5: Running OCR and Document Parsing Tasks with a Transformers BackendMay 18, 2026 · ProductLaunch → Granite Embedding Multilingual R2: Open Apache 2.0 Multilingual Embeddings with 32K Context — Best Sub-100M Retrieval QualityMay 14, 2026 · Research → Unlocking asynchronicity in continuous batchingMay 14, 2026 · ProductLaunch → v0.30.0May 12, 2026 · ProductUpdate → Building Blocks for Foundation Model Training and Inference on AWSMay 11, 2026 · Research → vLLM V0 to V1: Correctness Before Corrections in RLMay 6, 2026 · PolicyChange → Adding Benchmaxxer Repellant to the Open ASR LeaderboardMay 6, 2026 · ModelRelease → Granite 4.1 LLMs: How They’re BuiltApr 29, 2026 · OpenSourceRelease → DeepInfra on Hugging Face Inference Providers 🔥Apr 29, 2026 · PriceChange → Introducing NVIDIA Nemotron 3 Nano Omni: Long-Context Multimodal Intelligence for Documents, Audio and Video AgentsApr 28, 2026 · Research → How to build scalable web apps with OpenAI's Privacy FilterApr 27, 2026 · ProductUpdate → DeepSeek-V4: a million-token context that agents can actually useApr 24, 2026 · ProductLaunch → How to Use Transformers.js in a Chrome ExtensionApr 23, 2026 · ProductUpdate → QIMMA قِمّة ⛰: A Quality-First Arabic LLM LeaderboardApr 21, 2026 · Research →