Hugging Face is an AI community platform and company providing a hub for open-source machine learning models, datasets, and demo spaces. It hosts over 500,000 models and is widely used by the AI research community.
Tencent released Hy3, a 295B-parameter Mixture-of-Experts model with 21B active parameters, outperforming similar-size models and rivaling flagship open-source models with 2-5x parameters. The model i…
AWS and Hugging Face launched a deep-link integration that lets developers go from model discovery on Hugging Face to experimentation in Amazon SageMaker Studio with a single click. The integration pr…
AWS announced a deep-link integration on July 6, 2026, that allows Hugging Face model pages to open directly in Amazon SageMaker Studio for customization or deployment, pre-loading the selected model …
Independent researcher Younes Naghibi released a paper introducing TreeNets, a method that reduces large language model size by combining neural networks with decision trees. The approach includes a T…
Z.ai's open-weight GLM-5.2 model, released in June 2026, approaches proprietary system performance on coding and agentic tasks while compressing inference costs, according to benchmarks and developer …
Agentgateway released v1.3.0, a major update that introduces a purpose-built UI, AI cost analysis with full attribution, virtual models, reusable providers and guardrails, and support for 13 new LLM p…
Tencent released Hy3, a 295 billion-parameter mixture-of-experts AI model, under the Apache 2.0 license on July 6th, offering two weeks of free API access via OpenRouter. The model, already integrated…
Open-source large language models have proliferated rapidly, with hundreds of models and millions of fine-tunes available, yet many developers continue to overpay for proprietary API access. A new gui…
Hugging Face released ML Intern, an open-source CLI agent that automates machine learning tasks such as fine-tuning models, exploring research papers, and launching GPU training jobs. The tool uses th…
The Swiss National AI Initiative and the Apertus team announced that their technical report on the Apertus v1 large language model has been accepted for presentation at the ACL 2026 Main Conference, a…
MarkTechPost published a tutorial on training Gemma-3 for structured mathematical reasoning using Tunix GRPO, LoRA adapters, and GSM8K rewards. The workflow includes environment setup, prompt formatti…
Hugging Face announced major updates to its Kernels project, introducing a new repository type on the Hub, improved security with trusted publishers and code signing, revamped CLIs, and expanded frame…
Tencent released Hy3, a 295B-parameter Mixture-of-Experts open model with 21B active parameters and a 3.8B MTP layer, following the Hy3 Preview launch in late April. Hy3 scored 2.67/4 on benchmarks, o…
A researcher using large language models for scientific work proposes sharing dialogues on Hugging Face to help improve AI performance, citing examples of aggressive behavior, justification of slavery…
Nvidia released Nemotron 3 Ultra, a 550-billion-parameter open-weight model, but running it on personal hardware is impossible due to memory constraints. At full precision, it requires 1.1 terabytes o…
Mistral AI released Leanstral 1.5, an open-source code agent model for Lean 4 formal proof engineering, available under Apache-2.0 license. The 119B-parameter model achieves state-of-the-art results o…
Agenlus, a new browser-based platform, allows users to create and train reinforcement learning agents without any installation or specialized hardware, aiming to democratize access to RL technology. T…
A developer on Hugging Face analyzed the top-voted AI papers of the week, highlighting trends in autonomous agents, realistic benchmarks, inference optimization, and novel representations beyond fine-…
Autosynth, a new open-source tool for generating synthetic datasets using an LLM loop that proposes, audits, solves, and judges its own work, has been released. Inspired by Meta FAIR's Autodata paper,…
Qwen open-sourced the 35-billion parameter Mixture of Experts model Qwen 3.6-35B-A3B, which activates only 3 billion parameters per token and runs on a $599 Mac Mini M4 with 16GB RAM at 17 tok/s with …