redditPerplexity open-sourced their Mac inference server for Qwen 3.6redditlocal-llmLocalLLaMA3 days ago▲ 10Read story→
redditRunning 104GB Qwen3.8-Flash-Next on 48GB Mac at ~12 tok/sredditlocal-llmLocalLLaMA4 days ago▲ 10Read story→
YhackernewsShow HN: Running 104GB Qwen3.8-Flash-Next on 48GB Mac with at ~12 tok/shackernews5 days ago▲ 7Read story→
TMtechmemePerplexity launches Hybrid Compute, which splits workloads between frontier cloud models like Opus 5 and local LLMs, for all users of its Mac app (Igor Bonifacic/Engadget)industry5 days ago▲ 24Read story→