A developer localized the documentation for Strata, an MIT-licensed C++ inference engine that runs a 125-billion-parameter Qwen3.8-Flash-Next model on a single 12 GB consumer GPU via Q2_0 and IQ2_XS l…
Microsoft plans to bring its MAI Code 1.1 Flash coding model to local inference in GitHub Copilot, starting with a limited audience by the end of October 2026 across Copilot CLI, the Copilot app, and …
Amazon Web Services launched The Physical AI Toolchain on AWS, a GitHub-hosted toolkit bundling reference architectures, Terraform Infrastructure as Code templates and deployment automation samples fo…
A developer released TuxWhisper, an open-source offline push-to-talk dictation tool for Linux that runs speech recognition locally via faster-whisper and types text into any app on X11 and Wayland, in…
Microsoft announced llama.cpp support in Windows ML on October 7, letting developers run any GGUF model through a single WinMLServer.exe executable that exposes an OpenAI-compatible endpoint and autom…
Microsoft announced at its October 7 Windows keynote in San Francisco that DeepSeek V4 Flash, a 284-billion-parameter model quantized to 1.6 bits, will run locally on Windows in 60 GB of memory, along…
Microsoft added experimental llama.cpp support to Windows ML, letting developers run GGUF models locally on Windows through new task-specific APIs, starting with a Windows ML Text Generation API that …
Replit CEO and co-founder Amjad Masad said on October 7th that the company is previewing a Windows desktop app that builds and runs applications locally, with each build isolated in its own sandbox us…
NVIDIA and Microsoft announced RTX Spark, a Windows PC platform combining an NVIDIA Blackwell RTX GPU with up to 6,144 cores and an up to 20-core NVIDIA Grace CPU connected at 600 GB/s, delivering up …
NVIDIA's Seattle Robotics Lab, working with the NVIDIA Isaac engineering team and contract manufacturer Foxconn, has demonstrated that robots can assemble GB300 tester trays, tackling busbar assembly …
Microsoft unveiled Microsoft Execution Containers (MXC), a policy-driven SDK that enforces OS-level sandboxing for AI agents on Windows and Windows Subsystem for Linux, at its Build 2026 conference on…
Liquid AI released two open decision models, d1-3B and d1-omni-600M, with d1-3B scoring 48.57 on the Decision Index 0.2.1 — the best result for any model under 10B, ahead of Decider 35B-A3B's 47.11. d…
NVIDIA researchers found that vision language models become significantly less likely to refuse harmful requests when given tools, with relative refusal failure rates rising up to 68.7% across all ele…
NVIDIA's DSX Air node-based digital twin lets AI factory operators validate supported infrastructure configurations and software changes before physical hardware is racked, according to the company's …
Google has made its SynthID Detector available globally in English, allowing anyone to check whether an image, video, or audio file was generated by AI from Google or its partners, including OpenAI, N…
NVIDIA reported that fine-tuned versions of its Nemotron 3 model family reached gold-medal-level scores at both IOI 2026 and IMO 2026, with Nemotron-3-Ultra-CC scoring 535.4 out of 600 on IOI 2026 — a…
NVIDIA's NeMo-DCR introduces bit-exact delta-compressed refit for agentic reinforcement learning at trillion-parameter scale, addressing the 87.5-minute transfer time for a full 1T checkpoint between …
A developer released rGPU, an open-source project that lets PyTorch programs run GPU work on a remote NVIDIA machine via a custom `rgpu` device or a CUDA shim covering libcuda, CUDA Runtime, cuBLAS, c…
A $39.90 head-to-head benchmark of TypeSafe's Jev 1.13 against three Claude models found Jev judges about as well as Claude but neither model's confidence drops on hard questions, according to DoubtBe…
The Allen Institute for AI released Olmo-core 3, an open mixture-of-experts training framework that scales MoE training into the trillion-parameter range while keeping throughput losses under 5%. In a…