metal-shaders

1 article

sort: new top best

bug-bounty525 xss296 rce184 google175 exploit143 microsoft137 facebook137 malware135 account-takeover122 bragging-post117 cve113 privilege-escalation96 open-source88 csrf88 authentication-bypass83 phishing78 stored-xss75 access-control69 ai-agents66 apple64 web-security64 reflected-xss63 writeup63 reverse-engineering54 input-validation53 sql-injection51 ssrf51 browser50 cross-site-scripting49 dos48 smart-contract48 defi48 api-security47 supply-chain47 lfi45 ethereum45 information-disclosure44 tool43 privacy43 cloudflare41 web-application39 race-condition38 ctf38 opinion37 ai-security37 burp-suite37 vulnerability-disclosure37 llm37 web337 oauth36

0 2/10

Launch HN: RunAnywhere (YC W26) – Faster AI Inference on Apple Silicon

tool

RunAnywhere released MetalRT, a Metal GPU-optimized inference engine for Apple Silicon that achieves 1.67x faster LLM decode than llama.cpp and 4.6x faster speech-to-text than mlx-whisper through custom GPU shaders and zero-allocation inference. They also open-sourced RCLI, a voice AI pipeline combining STT, LLM, and TTS with sub-600ms end-to-end latency entirely on-device.

ai-inference apple-silicon gpu-optimization metal-shaders llm-performance speech-recognition text-to-speech on-device-ai macos open-source-tool

RunAnywhere MetalRT RCLI YC W26 Sanchit Shubham llama.cpp Apple MLX Ollama sherpa-onnx mlx-whisper Qwen3 LFM2.5

github.com · sanchitmonga22 · 3 days ago · details · hn