AI BriefingHuggingFace Blog news
The latest AI and developer news from HuggingFace Blog, with short summaries.
Visit website →PickUncensored Qwen3.8-27B MLX Model Released
An abliterated version of Qwen3.8-27B with censorship removed has been released in MLX format for Apple Silicon.
Agent Memory Optimization Strategies by Model Scale
Performance and costs can be optimized by adjusting the amount of guidelines provided to agents based on model performance and scale.
Sentence Transformers Adds Support for ColBERT Style
The Sentence Transformers v6.0 update adds support for ColBERT-style Multi-Vector embedding models.
Allocation Strategy Increases GPU Utilization by 33 Percentage Points
A constraint-based GPU allocation method improved GPU utilization by 33 percentage points compared to the existing FIFO method.
PickQwen 3.8 27B Released: Enhanced Coding and Agent Performance
The Qwen 3.8 27B model has been released, featuring significant improvements in coding, research, agent tasks, and vision-language understanding.
PickQwen3.8-27B Model Released
The next-generation open model Qwen3.8-27B, with enhanced coding and vision capabilities, has been released.
2026 Open Model Landscape Report
In the first half of 2026, the open model market is bifurcating into China's offensive with massive models and a US ecosystem centered on hardware companies.
Implementing a Robot Data Loop Based on Strands and LeRobot
Explains how to build a data loop from robot demonstration recording to training and deployment using Strands Robots and Hugging Face Storage Buckets.
Reproducing 2,200 ICML Papers Using AI Agents
A large-scale effort to reproduce over 2,200 ICML papers using AI agents verified the reliability of research.
PickQwen3.8 2.4T Released
Qwen has released the open model Qwen3.8 with a total of 2.4T parameters and 95B active parameters.
OlmoEarth Embedding Export Support
OlmoEarth Studio now supports on-demand generation and export of Earth observation data embeddings.
LFM2.5-VL-3B Enhances Edge Vision
Liquid AI released a 3.1B vision-language model with enhanced screen understanding and tool use.
PickNVIDIA Releases Magpie TTS
NVIDIA has released Magpie TTS, an open-weight model supporting 12 languages.
Meta Releases Muse Glimmer
Meta has released Muse Glimmer, a 30B multimodal model for local agents, under the Apache 2.0 license.
When Should AI Tutors Intervene?
TutorMoments evaluates AI tutors' excessive assistance and judgment capabilities using real classroom data.
Hugging Face Integrates Baseten
Hugging Face has added Baseten as an Inference Provider to support serverless model inference.
Ai2 Unveils OlmoEarth Platform
Ai2 has unveiled a geospatial foundation model and inference platform that processes large-scale satellite data.
Liquid AI Unveils CPU-Optimized LFM2.5 Encoder Models
Liquid AI has released two LFM2.5 encoder models capable of fast long-context processing even in CPU environments.
PickMoonshot AI Unveils Kimi K3 with 2.8T Parameters
Moonshot AI has released Kimi K3, an open-weight multimodal model with 2.8T parameters.