9 results for “local AI”
Hugging Face hired Jun Kim, creator of the oMLX library, to support the MLX ecosystem on Apple Silicon. The move funds a key bridge between Transformers and MLX, speeding local AI on Mac hardware.
NVIDIA released the PAIR beta on September 3, 2026, letting local AI agents route inference across RTX PCs, DGX Spark, and Macs on a home network. The router works with Ollama and LM Studio without agent changes.
Meta's Superintelligence Labs has released Muse Glimmer, a 30-billion-parameter open-weight model built specifically for always-on local agent workflows. Local agentic models change the economics of AI deployment.
A developer known as slvDev has demonstrated a 28.9-million-parameter language model generating text at 9.88 tokens per second on an ESP32-S3. Running a language model on a sub-$10 microcontroller changes the cost floor for local AI.
Tencent open-sourced an extreme quantized GGUF build of its 770B Hunyuan Hy4 model, compressing 1.5TB weights down to 214GB. By varying precision from 1.31-bit on non-critical layers to 2+ bits on sensitive layers, the model fits on accessible multi-GPU nodes.
Hugging Face released @huggingface/kernels with 207 versioned WebGPU kernels and the Fleet benchmarking tool for browser-based AI. The kernels beat ORT WebGPU by 2.57× geometric mean on Apple M4, creating a shared performance foundation for local inference.
Apple Intelligence has been registered with China's cyberspace regulator, clearing a major regulatory hurdle that could finally bring the company's AI. China is one of Apple's largest and fastest-growing markets.
Apple debuted its first 2 nm chip, the M6, in the new Mac mini and the quad-die M5 Ultra in the new Mac Studio. The M5 Ultra's 512 GB unified memory and 1.2 TB/s bandwidth let developers run massive AI models locally without cloud costs.
Anonymous model Ox Alpha launched on OpenRouter August 20 with a 1M-token context window and free preview access. Independent testing on a 10-task DeepSWE subset showed an 80% pass rate, beating Claude Fable 5, GLM-5.3, GPT-5.6 Sol, and Grok 4.6 on the same tasks.