Apple's Metal GPU glossary 'metalworking' targets developers familiar with CUDA but new to Apple Silicon, mapping every concept from GPU cores to MLX's architecture with CUDA equivalents and real kern…
Apple announced new Mac Studio and Mac mini models with the M6, its first 2nm chip for Macs, and the M5 Ultra, now the most powerful chip in the lineup, emphasizing local AI inference. The refresh fol…
Eigen Labs, a Seattle-based tech company, estimates that owners of Apple Silicon hardware can earn $120 to $200 per month by selling idle compute power for AI inference through its Darkbloom network, …
Developer Jon Ready released vllm-ios, a native Swift implementation of vLLM-style continuous batching for MLX on iOS, achieving 88% faster multi-agent inference than llama.cpp at 8 concurrent streams…
A developer built a capture-the-flag arena where language models attack and defend containers, and trained a local Qwen2.5-3B-Instruct bot with an MLX LoRA adapter on game replays. The bot outperforme…
Local-first workspace software can eliminate the need for cloud indexing by leveraging on-device AI and search capabilities, according to a guide from Neutron Tech. Modern hardware such as Apple Silic…
Base Compute, an AI inference company operating in Melbourne and Berlin, released results on August 20 from its Base Optimization Stack (B:OS), a system that uses AI agents to port and tune open-weigh…
Ollama v0.32.0, released July 11, changed the bare `ollama` command from printing help text to launching a local coding agent, pivoting the project from model runner to agent runtime. The update intro…
LM Studio's local LLM performance hinges on whether users run MLX on Apple Silicon Macs or GGUF on Windows GPU laptops, with Apple's unified memory enabling larger models while discrete GPUs like the …
Alibaba's Qwen 3 27B, a dense open-weight multimodal model, can be run locally using DeepSeek Harness, achieving agentic performance close to Claude 4.5 on benchmarks. On a two-node NVIDIA DGX Spark c…
As of August 2026, Ollama leads local LLM runtimes with roughly 179,000 GitHub stars, offering an MIT-licensed tool with a desktop GUI and OpenAI-compatible endpoint, while LM Studio is recommended fo…
A developer demonstrates how to run fine-tuned large language models on Apple Silicon using MLX Swift bindings, bypassing CoreML's compilation latency. The approach uses lazy evaluation, direct loadin…
Apertura, a from-scratch Objective-C++/MLX rebuild of Google's Gemma-4 for Apple Silicon, runs the entire Gemma-4 text family natively on a Mac with bit-exact conformance to the PyTorch reference, ach…
Qwen3.8-27B, a 27-billion-parameter dense model from Alibaba, ranks #1 of 135 models on Artificial Analysis's Intelligence Index with a score of 52, one point above GLM-5.2, a 753-billion-parameter op…
Llmfit, a Rust CLI and TUI created by Alex Jones, the developer behind k8sgpt, right-sizes local LLMs to a user's hardware by detecting RAM, CPU, and GPU and scoring hundreds of models on quality, spe…
Ramabana, a new coding agent from developer Vedic Reader, is now available via pip, featuring a host that coordinates policy, tools, memory, and routing, with models loaded through the rishi tool supp…
A developer has released a clone-and-run launcher that runs LTX-2.5 text-to-video generation locally on Apple Silicon in a single command, with audio support and no need for ComfyUI or CUDA. The launc…
Liquid AI released LFM2.5-VL-3B, a 3.1B-parameter vision-language model for on-device deployment, averaging 69.4 across 28 vision benchmarks, matching InternVL-3.5-4B and 0.7 points behind Qwen3.5-4B.…
A developer has created a privacy-preserving health AI using Apple's MLX framework and Meta's Llama-3 model, enabling 100% offline analysis of Apple Health data on MacBooks. The approach leverages App…
A developer fine-tuned Mistral 7B on a laptop to detect personal data in log lines and support messages, achieving a perfect 100% score on an initial test set. However, after rebuilding the test set f…