Perplexity releases pplx-decider-flash-v1.2 for local Mac execution from the menu bar
A Reddit post highlights Perplexity's pplx-decider-v1.2 and pplx-decider-flash-v1.2, open-source models released under Apache 2.0 and available on Hugging Face. The poster claims pplx-decider-v1.2 costs one tenth as much as GPT 6 Luna (low reasoning) while producing identical-quality decisions, citing Artificial Analysis' Decision Index. If the pricing claim holds, decision-style API workloads become far cheaper to run.
A Reddit post highlights Perplexity's pplx-decider-v1.2 and pplx-decider-flash-v1.2, open-source models released under Apache 2.0 and available on Hugging Face. The poster claims pplx-decider-v1.2 costs one tenth as much as GPT 6 Luna (low reasoning) while producing identical-quality decisions, citing Artificial Analysis' Decision Index. If the pricing claim holds, decision-style API workloads become far cheaper to run.
Perplexity open-sourced its Mac inference server, released under the 'lily' directory of its pplx-garden GitHub repository. The server is optimized for running a single model to get the best performance on Apple Silicon. It gives local Mac developers a purpose-built serving stack tuned for Qwen 3.6.
Perplexity open-sourced its Mac inference server, released under the 'lily' directory of its pplx-garden GitHub repository. The server is optimized for running a single model to get the best performance on Apple Silicon. It gives local Mac developers a purpose-built serving stack tuned for Qwen 3.6.
Nativ is a tool that enables running frontier open-weight AI models locally on Mac computers. It aims to bring high-performance inference to consumer hardware without cloud dependencies.
Nativ is a tool that enables running frontier open-weight AI models locally on Mac computers. It aims to bring high-performance inference to consumer hardware without cloud dependencies.
A free Claude Code skill that helps website owners optimize their pages to be cited by AI answer engines like ChatGPT, Perplexity, and Google AI Overviews. It was tested on a low-authority Amazon affiliate site, where AI engines became the top traffic source. The skill likely analyzes page structure and suggests improvements to make content more quotable.
A free Claude Code skill that helps website owners optimize their pages to be cited by AI answer engines like ChatGPT, Perplexity, and Google AI Overviews. It was tested on a low-authority Amazon affiliate site, where AI engines became the top traffic source. The skill likely analyzes page structure and suggests improvements to make content more quotable.
An open source tool that lets an existing MLX model running locally on a Mac act as a decider without additional fine-tuning. The LLM makes accept/reject or choice decisions on inputs directly. It targets developers running local inference on Apple silicon who want decision outputs without training overhead.
An open source tool that lets an existing MLX model running locally on a Mac act as a decider without additional fine-tuning. The LLM makes accept/reject or choice decisions on inputs directly. It targets developers running local inference on Apple silicon who want decision outputs without training overhead.