Unsloth Desktop, a new local AI training tool from Unsloth, enables users to train models locally without a cloud backend, supporting NVIDIA, AMD, Intel, and Mac hardware. It claims a 70% reduction in…
Cua, a company building computer-use agent infrastructure on macOS virtual machines, achieved 11–16× faster llama.cpp inference inside Apple Silicon VMs by intercepting Metal capability queries to rep…
Meta released Muse Glimmer, a 30-billion-parameter AI model optimized for always-on local agent workflows, available under the Apache 2.0 license. The model supports tool calling, function calling, lo…
LiquidAI and collaborators released fuse-1 Lite, a 5.72 billion parameter mixture-of-experts model that transplants 960 coding experts from Qwen3.6-35B-A3B into LiquidAI's LFM2.5-2.6B host model, achi…
Fuse-1 Lite, a 5.72B parameter mixture-of-experts coding model from LiquidAI, can run locally with VRAM needs ranging from 3.36 GB in 4-bit quantized form to about 12 GB in full bfloat16 precision, ac…
Meta Superintelligence Labs released Muse Glimmer, a 30B multimodal open model for local agent workloads, now available on Ollama with initial support for Apple Silicon via the MLX engine. The model, …
Meta's Superintelligence Labs released Muse Glimmer, a 30-billion-parameter dense multimodal model with 131K context, under the Apache 2.0 license, marking a departure from the restrictive Llama Commu…
Meta released Muse Glimmer, a 30-billion-parameter AI model that runs on a single consumer GPU without cloud connectivity, on August 10. The model, compressed to under 20 GB via logit distillation, qu…
Meta Superintelligence Labs released Muse Glimmer, a 30-billion-parameter open-weights model optimized for local agent workflows, under an Apache 2.0 license. The model runs on a single consumer GPU a…
FastVideo released FastMetal-QAD, a family of three open-source video generation models (1.3B, 5B, and 14B) that run natively on Apple Silicon via a new MLX runtime, with the 5B model generating a fiv…
Meta Superintelligence Labs released Muse Glimmer, a 30-billion-parameter multimodal open model under Apache 2.0, designed for local agent workloads on consumer hardware. The model scores 76 on SWE-Be…
PipeNetwork released minimax-h3-mlx, a community MLX implementation of MiniMax-H3 for Apple Silicon, translating the 33-billion-parameter video-and-audio diffusion pipeline into native MLX. The reposi…
Simon Willison successfully ran MiniMax-H3, a text-to-video model, on an Apple M5 Max MacBook Pro using the PipeNetwork/minimax-h3-mlx Python package, which ports the model to MLX. The process downloa…
A developer built a private AI health assistant that runs entirely on a MacBook using Apple's MLX framework and Meta's Llama-3 model, eliminating the need to send sensitive health data to cloud server…
A developer reports that multi-agent collaboration is hitting an engineering wall as single-agent systems plateau, citing Mano AFK's experience where splitting coding and testing into separate context…
Developer ansnadeem released mlxsh, a lightweight command-line tool that serves multiple local MLX models on Apple silicon at OpenAI-compatible endpoints, supporting text-only and multimodal modes, an…
Symbio, a local AI assistant that learns from user corrections and fine-tunes itself without cloud or subscriptions, was released on GitHub by developer HuyEdits. The system runs on Apple Silicon usin…
On July 30, an open-source project ran Kimi K3, a 2.78-trillion-parameter model that requires 1.42 TB of storage, on a MacBook Pro with 64 GB of RAM at 0.3 tokens per second, demonstrating that memory…
Researchers at UC Berkeley Sky Lab extended the K-Search evolutionary kernel optimization framework with a CUDA-to-MLX translation layer, enabling automatic transfer of GPU kernel expertise from NVIDI…
LocalAI, a self-hosted runtime that runs AI workloads on user-controlled hardware, offers an OpenAI-compatible API supporting text generation, vision, speech, image and video generation, embeddings, a…