📦 Open Source

83 articles

🟢 📦 Open Source July 10, 2026 · 2 min read

PyTorch: 'free normalization' fuses Layer Norm into GEMM and attention kernels — Meta targets lower training cost for large models

The PyTorch team at Meta published techniques for fusing normalization operations directly into GEMM and attention kernels, with the goal of eliminating their computational cost — 'free normalization'. The code is available on GitHub through a kernel library, and the direct result is faster training and inference for models with frequent Layer Norm and RMS Norm operations.

🟡 📦 Open Source July 9, 2026 · 2 min read

Ollama: $88 million in funding for local AI — 8.9 million developers and Docker's founder among investors

Ollama is a platform for running open AI models locally that announced $88 million in funding with investors including Benchmark, Theory Ventures, and Solomon Hykes, Docker's founder. The platform serves 8.9 million developers and 85% of Fortune 500 companies, while its cloud service doubles token volume every month.

🔴 📦 Open Source July 8, 2026 · 5 min read

PyTorch 2.13: Up to 4× Less GPU Memory for LLM Training and 12.3× Faster FlexAttention

PyTorch 2.13 ships with 3,328 commits from 526 contributors. Key highlights: nn.LinearCrossEntropyLoss cuts peak GPU memory footprint by up to 4× for LLM training, FlexAttention on Apple Silicon achieves up to 12.3× speedup, and the new torchcomms backend modernizes distributed training.

🟡 📦 Open Source July 8, 2026 · 4 min read

CNCF White Paper: Data Storage Remains the Main Barrier to Cloud-Native AI at Scale

The CNCF TAG Infrastructure published a white paper mapping data storage bottlenecks in cloud-native AI environments, distinguishing the requirements of training, inference, and agentic AI phases, and providing architectural guidelines for Kubernetes AI/ML deployments.

🟢 📦 Open Source July 8, 2026 · 3 min read

IBM and Red Hat Expand Lightwell — 6,500+ Remediated Open-Source Dependencies and Clearinghouse for Finance

IBM and Red Hat launched Lightwell Network in general availability with a catalog of 6,500+ digitally signed remediated dependencies in Java and Python, plus Lightwell Clearinghouse Premier for coordinating patch embargos in financial services, backed by a $5 billion commitment to open-source security.

🟡 📦 Open Source July 6, 2026 · 4 min read

vLLM and Tencent Hunyuan upstream two HPC backends for NVIDIA Hopper GPUs

The Tencent Hunyuan AI Infra team and the vLLM team have jointly upstreamed the HPC_ATTN attention backend and hpc MoE backend, reducing TTFT by 24% and TPOT by 17% on an 8x NVIDIA H20 configuration, without the need to fork vLLM code.

🟡 📦 Open Source July 1, 2026 · 4 min read

Kimi K2.7 Code Arrives in GitHub Copilot as the First Open-Weight Model

Moonshot AI's Kimi K2.7 Code is available in GitHub Copilot as the first open-weight model in the model picker, hosted on Microsoft Azure. It is available on Pro, Pro+, and Max plans with usage-based billing, with planned expansion to Business and Enterprise.

🟢 📦 Open Source July 1, 2026 · 3 min read

LangChain Releases OpenWiki: Open-Source Tool That Automatically Documents Repositories for AI Agents

LangChain has released OpenWiki, an open-source CLI tool that automatically generates and updates documentation wikis for code repositories. The goal is to give coding agents structured context about a repository without bloating instruction files like AGENTS.md or CLAUDE.md. The tool is built on the DeepAgents framework and supports multiple LLM providers.

🟡 📦 Open Source June 30, 2026 · 4 min read

Miles: A PyTorch-Native Open-Source Framework for RL Post-Training of Frontier-Scale LLMs

RadixArk releases Miles, an open-source reinforcement learning framework that integrates SGLang, Megatron-LM, Ray, and PyTorch into a single production-tested stack for post-training large language models on Hopper and Blackwell GPUs.

🟢 📦 Open Source June 30, 2026 · 4 min read

CNCF Kepler Rearchitected From the Ground Up: Precise Pod Energy Measurement Without Kernel Privileges

CNCF sandbox project Kepler has been completely rewritten: the new architecture replaces the eBPF approach by reading standard /proc and /sys paths, eliminates multi-kilowatt measurement spikes, and reduces the process power attribution gap to the milliwatt level.

🟢 📦 Open Source June 30, 2026 · 3 min read

ONNX v1.22.0 Brings Native Attention Operators for LLMs and WebAssembly Support

LF AI & Data Foundation has released ONNX v1.22.0 with native attention operators for transformer architectures and LLMs, WebAssembly support for in-browser model inspection, and SLSA Level 2 cryptographic attestations. Twenty-seven contributors participated, 16 of them for the first time.

🟡 📦 Open Source June 29, 2026 · 2 min read

NVIDIA: Palantir and NVIDIA Nemotron Bring Sovereign AI to US Agencies in Air-Gapped Systems

Palantir has integrated NVIDIA Nemotron open models into its Sovereign AI Operating System for US government agencies and critical infrastructure. The system operates in air-gapped environments — completely isolated networks with no internet connection — covering defense, energy, healthcare, education, and transportation sectors.

🟡 📦 Open Source June 27, 2026 · 2 min read

GitHub: Desktop 3.6 brings Git worktrees and AI merge conflict resolution

GitHub Desktop 3.6 is available for macOS and Windows, introducing support for Git worktrees — a mechanism that enables parallel work on multiple branches without stashing — along with deeper Copilot integrations for commit messages and merge conflicts.

🔴 📦 Open Source June 26, 2026 · 2 min read

PyTorch: TokenSpeed-Kernel — portable high-performance kernels for multi-silicon LLM inference

TokenSpeed-Kernel is an open-source three-layer kernel subsystem that accelerates LLM inference on NVIDIA and AMD GPUs by up to 3.6× without code rewrites, and is already integrated into the vLLM inference framework.

🟢 📦 Open Source June 26, 2026 · 1 min read

Allen Institute: which tokens does a hybrid model (OLMo 3) predict better?

The Allen Institute (AI2) analyzes OLMo 3 and OLMo Hybrid architectures, revealing that hybrid models better predict semantic, context-dependent tokens, while pure Transformers remain superior for verbatim text copying.

🟢 📦 Open Source June 25, 2026 · 1 min read

arXiv:2606.24855: OpenThoughts-Agent — open data recipes for training agentic models

OpenThoughts-Agent is an open pipeline for curating data for agentic language models. Through 100+ ablation experiments, the team built 100K examples and fine-tuned Qwen3-32B, achieving 44.8% on seven agentic benchmarks — surpassing all existing open-source models.

🟢 📦 Open Source June 21, 2026 · 2 min read

Anthropic (Claude Code GitHub): Claude Code v2.1.185 improves stream stall message

Claude Code v2.1.185 is a UX patch released on June 20, 2026 that changes the API stream stall message and extends the activation threshold from 10 to 20 seconds of silence — with no changes to the model or API.

🟡 📦 Open Source June 20, 2026 · 2 min read

arXiv:2606.20517: Multi-LCB Extends LiveCodeBench to 12 Programming Languages and Reveals Python Overfitting in 24 Models

Multi-LCB is an extension of the LiveCodeBench benchmark from Python to 12 programming languages, described in the paper arXiv:2606.20517 and accepted at ICLR 2026. By testing 24 large language models, the authors reveal significant Python overfitting and language-specific data contamination, directly exposing the limits of multilingual code generation in today's models.

🟡 📦 Open Source June 20, 2026 · 2 min read

UK AISI: Engineering Playbook Opens Frontier Model Evaluation Infrastructure in Five Layers

The Engineering Playbook is open-source documentation published by the UK AI Safety Institute on June 18, 2026 that opens up the internal infrastructure for evaluating frontier AI models. The Playbook is structured in five layers (Evaluate, Isolate, Connect, Run, Scale) and builds on the previously released Inspect AI tool, which has over 200 ready-made evaluations and 240 contributors.

🟡 📦 Open Source June 19, 2026 · 2 min read

Black Forest Labs: Robin Rombach Calls on G7 Leaders to Support Open AI Development

Robin Rombach, co-founder and CEO of Black Forest Labs (creators of the FLUX models), addressed G7 leaders with a call for open and responsible AI development to become a global norm. Rombach argues that public availability of AI parameters drives innovation and democratizes technology.

🟡 📦 Open Source June 18, 2026 · 2 min read

Allen Institute: Open-Source MolmoMotion Predicts 3D Motion from Video and Sets SOTA in Robotics

Allen Institute released MolmoMotion, a fully open-source model that predicts 3D object trajectories from video and natural language instructions such as 'rotate the bowl'. The model achieves state-of-the-art on PointMotionBench with 0.109 m average displacement versus 0.134 m for the previous record, and raises pick-and-place task success in robotics from 56% to 76.3%, a gain of 20.3 percentage points. It was trained on the MolmoMotion-1M dataset of 1.16 million videos with 3D trajectories and action descriptions.

🟡 📦 Open Source June 17, 2026 · 1 min read

vLLM: Semantic Router Fusion Combines a Model Panel with a Judge That Synthesizes a Single Response

vLLM introduced Semantic Router Fusion, a primitive in which multiple models work in parallel as a panel, and a judge model analyzes consensus and differences to synthesize a single response. It supports local vLLM and private endpoints as well as public providers such as Gemini, Kimi, DeepSeek, and Claude. External validation on OpenRouter DRACO showed 69% for the fused panel versus 65.3% for the best single model, with full OpenAI API compatibility.

🟡 📦 Open Source June 16, 2026 · 2 min read

GitHub: Open Multilingual Repositories Dataset with 80 Million Rows and 40 Million Repositories

GitHub has published the Multilingual Repositories Dataset with more than 80 million classification rows across 40+ million repositories, under the fully open CC0-1.0 license. For each repository the dataset records three text sources — the README, the most-commented issue, and the most-commented pull request — alongside language detection through three tools: fastText, gcld3, and lingua-py. Portuguese leads among non-English README files, while Korean is the most represented in issue discussions.

🟢 📦 Open Source June 16, 2026 · 2 min read

CNCF: Oracle's $3 Million Donation in OCI Credits Accelerates Arm64 Support in 12+ Projects

CNCF reports that Oracle's donation of $3 million in OCI compute credits enables Arm64 CI/CD support in more than 12 projects, including OpenTelemetry, containerd, Falco, Longhorn, Crossplane, and Jaeger. Demand quickly outpaced the initial guideline of $5,000 per month. The shift follows data showing that more than 50 percent of new AWS instances and 33 percent of Azure instances now run on Arm64 architecture.

🟡 📦 Open Source June 12, 2026 · 4 min read

Allen Institute for AI Released olmo-eval: Evaluation Platform Designed for the Active LLM Development Cycle

AI2 has released olmo-eval, an open evaluation platform that extends the OLMES standard from one-shot benchmarking toward a full development loop. The key innovation is per-question pair comparison across checkpoints, with built-in statistical measures for distinguishing genuine improvements from data noise.

🟡 📦 Open Source June 12, 2026 · 4 min read

MiniMax M3 in vLLM: Day-0 Support for a Model with 1-Million-Token Context

The vLLM team has released full Day-0 support for MiniMax M3 — a model with 1-million-token context, native multimodality, and a Mixture-of-Experts architecture. The key innovation is MiniMax Sparse Attention, which makes practical long-context serving possible without fully materializing the KV cache.

🟡 📦 Open Source June 11, 2026 · 4 min read

Ollama delivers up to 20% faster Apple Silicon inference: MLX engine, NVFP4 quantization, and snapshots for agentic workflows

Ollama has updated its MLX engine for Apple Silicon, delivering up to 20 percent faster token generation through fused Metal kernel operations and GPU sampling. The new NVFP4 quantization format halves quality loss compared to unquantized BF16, and a snapshot system simplifies agentic workflows with branching and retry support.

🟢 📦 Open Source June 10, 2026 · 3 min read

Helion in vLLM: PyTorch-native DSL replaces CUDA kernels and accelerates FP8 inference for Qwen3 models

Helion is a PyTorch-native DSL for writing GPU kernels without direct CUDA code. Nine Helion kernels have been integrated into vLLM for FP8 inference of Qwen3 models: non-GEMM kernels show speedups of up to 73% on the H100, and overall throughput grows by approximately 5–9% compared to previous solutions.

🟢 📦 Open Source June 9, 2026 · 4 min read

LF AI & Data: DocLang working group develops open standard for AI-native documents

LF AI & Data Foundation has launched the DocLang Specification Working Group with founding members IBM, Red Hat, and NVIDIA to develop an open standard for AI-native documents. The format preserves semantic meaning and geometric layout, embeds governance controls, and is optimized for modern AI tokenizers and agentic workflows.

🟢 📦 Open Source June 9, 2026 · 4 min read

vLLM: vime — new RL framework that unifies Megatron training and vLLM inference in a single pipeline

vime is a new RL framework for post-training language models that unifies Megatron distributed training and vLLM inference in a single pipeline. On GB200 hardware, the Qwen3-30B-A3B MoE model achieves approximately 147 seconds per step — 1.72x faster than on H200. R3 routing replay reduces the log-probability gap from 0.019 to 0.013.

🟡 📦 Open Source June 6, 2026 · 3 min read

arXiv:2606.08094: vla.cpp runs Vision-Language-Action models on 1.3 GiB of memory

A new paper presents vla.cpp, a C++ inference engine for running Vision-Language-Action policies on resource-constrained robotic hardware. The engine reaches SOTA level on the LIBERO-Object benchmark and runs BitVLA with just 1.3 GiB of memory.

🟡 📦 Open Source June 5, 2026 · 3 min read

Ollama 0.30: llama.cpp Integration, GGUF Support, and Up to 20% Faster Inference

Ollama 0.30 brings integration with llama.cpp for better performance and GGUF model compatibility, with up to 20% faster throughput on NVIDIA GPUs. It expands hardware support with Vulkan on AMD and Intel devices and adds support for tool-calling. It complements the existing MLX engine for Apple silicon.

🟡 📦 Open Source June 5, 2026 · 3 min read

vLLM Semantic Router v0.3 'Themis': Production-Grade Stateful Query Routing

The vLLM team has released v0.3 'Themis' of its Semantic Router, the first production-ready version for routing queries between models. It brings canonical configuration, an inspectable decision flow, and reproducible routing behavior for Kubernetes deployments.

🟢 📦 Open Source June 3, 2026 · 3 min read

Google open-sources the hydrology framework behind the Flood Hub flood-forecasting service

On 3 June 2026, Google Research released on GitHub under the Apache 2.0 license the Python/PyTorch framework that powers the global Flood Hub flood-forecasting service. The framework uses LSTM and the newer ME-LSTM architecture over the open-source Caravan dataset, and the upgraded v2 model extends the reliable forecast horizon by 6 days in gauged basins.

🟡 📦 Open Source June 2, 2026 · 2 min read

vLLM: Session-Aware Agentic Routing cuts model switches by 79 percent

The vLLM Semantic Router has gained Session-Aware Agentic Routing (SAAR), a mechanism that understands long-running agentic conversations instead of treating each message separately. Across 21,600 test turns, SAAR cuts model switches by 79.29 percent, fully eliminates unsafe switches, and lowers estimated costs by 78.71 percent.

🟢 📦 Open Source June 2, 2026 · 2 min read

vLLM: AutoRound quantization arrives in vLLM-Omni for smaller multimodal models

vLLM has integrated Intel's AutoRound quantization into vLLM-Omni, enabling W4A16 compression of multimodal and diffusion models. The result is up to a 62 percent smaller checkpoint, with minimal quality loss and faster generation on Intel XPU and NVIDIA graphics cards.

🟡 📦 Open Source June 1, 2026 · 2 min read

vLLM: running on NVIDIA DGX Spark / GB10 systems

The vLLM team published a practical guide for running vLLM on the NVIDIA DGX Spark system based on the GB10 chip. The guide covers unified memory behavior, serving the NVFP4 Nemotron-3-Super model, Docker deployment, Prometheus metrics and local evaluation results on the new edge hardware.

🟢 📦 Open Source May 30, 2026 · 2 min read

xAI: xai-sdk-python v1.15.0 adds the grok-build-0.1 model and context compaction

xAI has released xai-sdk-python v1.15.0, a release with three changes: adding the grok-build-0.1 model to the ChatModel class, support for context compaction, and bumping the version from 1.14.0 to 1.15.0. All changes arrived through three separate pull requests.

🟡 📦 Open Source May 29, 2026 · 2 min read

arXiv:2605.31463: PithTrain, a Compact MoE Training System Tailored to Agents

PithTrain is a new training system for Mixture-of-Experts models optimized for AI coding agents. The paper introduces the agent-task efficiency metric and the accompanying ATE-Bench, on which PithTrain achieves up to 62 percent fewer agent steps and 64 percent less active GPU time at comparable throughput.

🟡 📦 Open Source May 28, 2026 · 2 min read

vLLM: Standardized weight-synchronization APIs accelerate asynchronous RL

On May 28, 2026, vLLM introduced standardized APIs for weight synchronization and improved support for asynchronous reinforcement learning. It addresses the fragmentation in which every RL framework ad-hoc implements weight transfer between training and inference, and fixes stability problems in large distributed deployments.

🟢 📦 Open Source May 28, 2026 · 2 min read

Microsoft: Data Formulator 0.7 connects fragmented enterprise data

Data Formulator 0.7 is an open-source system from Microsoft Research that connects fragmented data sources in enterprises and supports iterative analytical workflows. It brings persistent data connectors and context-aware AI agents, enabling analytics without knowledge of SQL or programming.

🟢 📦 Open Source May 28, 2026 · 2 min read

Mistral: Search Toolkit for production search pipelines

Mistral Search Toolkit is an open-source framework for building production search pipelines for AI applications. It unifies ingestion, retrieval and evaluation under a single interface and can be deployed in the cloud, on-premises or at the network edge.

🟢 📦 Open Source May 28, 2026 · 2 min read

Ollama: OpenJarvis v1.0 for personal AI agents that run locally

OpenJarvis v1.0 is an open-source framework for building personal AI agents that run locally on your own hardware, with built-in Ollama integration. It is a local-first approach in which models stay local and the cloud is optional, with tracking of energy consumption, cost and latency.

🟡 📦 Open Source May 27, 2026 · 3 min read

PyTorch: How torch.compile achieves up to 10× speedup through kernel fusion

The PyTorch compiler torch.compile achieves up to 10× speedup on GPUs by eliminating redundant writes to global memory through kernel fusion. Instead of separate GPU kernels for each operation, the Inductor compiler automatically merges dependent operations into a single kernel that keeps intermediate values in the processor's fast registers.

🟢 📦 Open Source May 27, 2026 · 2 min read

xAI: SDK Python v1.14.0 adds enable_image_search to web_search and SERVER_SIDE_TOOL_IMAGE_SEARCH enum

xAI released Python SDK v1.14.0 which brings support for image search within the web_search functionality. The new enable_image_search field and SERVER_SIDE_TOOL_IMAGE_SEARCH enum constant allow development teams building Grok-based applications to enable server-side visual content search alongside standard web text search.

🟢 📦 Open Source May 23, 2026 · 4 min read

Kedro: version 1.2.0 brings the @experimental decorator and a LangGraph agentic starter for GenAI pipelines

Linux Foundation AI project Kedro released version 1.2.0 along with Kedro-Viz 12.3.0. The new @experimental decorator enables marking APIs under development, and the starter project support-agent-langgraph demonstrates integration with LangGraph orchestration and Langfuse/Opik prompt management tools. Kedro-Viz gains Mermaid diagrams and node preview extensibility for improved pipeline debugging.

🟡 📦 Open Source May 21, 2026 · 2 min read

Stability AI: Stable Audio 3.0 with open-weight models and generation up to 6 minutes

Stability AI released Stable Audio 3.0 on 20 May 2026 — a family of 4 audio models (Small SFX, Small, Medium, Large) of which three are open-weight and available on Hugging Face. The key advance is generation of audio up to 6 minutes (the previous version produced 47 seconds), along with support for audio inpainting, causal continuation, and LoRA fine-tuning. All models are claimed to have been trained exclusively on licensed data.

🟢 📦 Open Source May 20, 2026 · 2 min read

LangChain: The agent that fixes agents — how LangSmith Engine was built

LangChain published a detailed technical overview of LangSmith Engine — an autonomous agent that analyzes errors in production AI agents and proposes concrete fixes. It compresses thousands of traces, classifies them with a screener sub-agent, and generates validated evaluators for the Issue Board.

🟡 📦 Open Source May 19, 2026 · 2 min read

PyTorch: ExecuTorch MLX Delegate delivers 3–6× faster model execution on Apple Silicon GPUs

The PyTorch team released the experimental ExecuTorch MLX Delegate — a backend that leverages the Apple MLX framework and Metal GPU kernels for 3 to 6 times greater throughput on Apple Silicon chips. Supports Llama 3.2, Qwen 3, Phi-4 mini, Whisper and Voxtral real-time streaming transcription.

🟢 📦 Open Source May 19, 2026 · 2 min read

GitHub: Copilot Spaces API now generally available

GitHub announced the general availability of the REST API for Copilot Spaces, allowing teams to programmatically create, configure and delete contextual AI workspaces. The new interface is especially useful for organizations managing large numbers of Spaces without relying on manual workflows.