Open models, frameworks and repositories: weight releases, community projects that take off, and open versus closed.
Meituan has released the weights for LongCat-2.5, a 1.6-trillion-parameter model built for AI agents and trained entirely on Chinese hardware. On Terminal Bench, it scores 88.4 points, landing just behind Claude Opus 5. The sample deployment runs on 194 Huawei Ascend chips, and Meituan makes no mention of Nvidia GPUs.
llama.cpp makes it possible to run multi-modal LLMs across a wide range of hardware - from consumer-grade local machines to high-end enterprise servers. Today we release libraries v1.0.0, highlighting our responsibility to those developers.
LLMs rely on their context to generate the next set of tokens. To address this, we developed LLMON (LLM Object Notation, pronounced “lemon”), an LLM-native markup language for explicitly communicating structure and semantics to the models.
AI agents now run multi-step workloads for hours and act on behalf of users, raising the bar for secure, scalable agent compute. Learn how the next generation of Amazon Bedrock AgentCore runtime delivers hardware-isolated sessions with faster starts, more compute, usage-based billing, and isolation that now extends to your data.
Jensen Huang’s resignation raises questions about Nvidia’s commitment to open source AI, including its Hugging Face acquisition, open-weight models, and advocacy against restrictions, as the company navigates a leadership transition.
The Universal Agent Evaluation Framework (UAEF) is a framework-agnostic system for evaluating AI agents. UAEF converts each framework's traces into one canonical format, scores them with 36 built-in metrics across 7 dimensions, and stores results so teams can compare agent versions and catch regressions.
Kubernetes, the open source system that manages how applications run across thousands of AI factories, has become the foundation for AI, and the demands on it are only growing. To help meet them, NVIDIA is deepening its commitment to open source at KubeCon + CloudNativeCon, the flagship conference for the Kubernetes and cloud-native communities.
Simulation, which has long been the proving ground for robots learning to walk, roll and hop, is now helping researchers teach robots tasks that require a finer touch: aligning gears, connecting cables and assembling small components.
Today we're introducing a framework for creating these environments; we call them FoundationWorlds. Our first foundation world is ToyBlocks : an open simulation framework built around a robot arm and blocks on a tabletop.
Inside Inkling 2 Small’s training run and METR’s study of AI research at Anthropic.
NVIDIA Cosmos world model development platform adds four new open-weight checkpoints to its Cosmos 3 model family for generating interactive world simulation and multi-sensor multi-view driving scenes. All four models are post-trained examples of Cosmos 3 Nano and have 16B trainable parameters. The models use OpenMDW1.1, permitting commercial and non-commercial use under its terms.
Two weeks ago, we opened up Phonon, our small text-to-speech model that runs on your own device. It needs no server and no network, and what you type stays with you.
In production AI agents, context management is the real bottleneck. This post presents a four-layer context engineering framework (architecture, prompts, file-based tools, and verification agents), implemented in working code from Deep Insight, an open source multi-agent data analysis solution on the Strands Agents SDK and Amazon Bedrock AgentCore.
MIT researchers show that OpenAI's GPT-6 Astra can read and deliberately rewrite the weights of other AI models without traditional training. In one experiment, Astra boosted a Qwen model's math accuracy from 7.9 to 48.1 percent by changing just 0.016 percent of its weights. The approach could fundamentally simplify AI optimization but still hits clear limits.
An IBM team compared CUGA, Hermes and Deep Agents on 90 AppWorld tasks using the same open-weight model, GLM-5.3-flash. Task pass rates were 81.1%, 46.7% and 30.0%, and full-scenario pass rates were 66.7%, 26.7% and 6.7%.
Mistral says its new model can afford organizations a kind of control over its cyber defenses that closed models can't.
Mistral Large 4 (aka Le Chonk) is a multimodal, multilingual, Mixture-of-Experts (MoE) model from Mistral. It takes text and images as input and generates text. It has about 1T total, 49B active parameter, and a context window of 1M tokens.
The weights of Mistral Large 4 (Le Chonk) are now open. A 1T parameter multimodal model, and best open-weight model for coding, trained in Europe.
One open model for multilingual transcription, speaker attribution, translation, and audio Q&A, with top leaderboard results, plus the open Granary v2 data pipeline and fine-tuning recipes to adapt it.
LFM3 is our new generation of Liquid Foundation Models (LFMs), built to be the starting point for specialized intelligence. It brings text, vision, audio, omni, and decision models onto one shared backbone, all in the 2B parameter range.