Llama.cpp on a jailbroken PS5: run any GGUF model natively on the console

	Llama.cpp on a jailbroken PS5: run any GGUF model natively on the console(github.com)
	367 points by Anon84 9 days ago \| 84 comments
	tl;dr: PS5LM turns a jailbroken PS5 (firmware 7.00–13.60) into a private LLM server, running llama.cpp on the console's RDNA 2 GPU via Vulkan with ~11.4 GB of shared memory for GGUF models. It exposes OpenAI and Anthropic-compatible APIs over the network, plus a /v1/systemone endpoint for decision models, hitting roughly 55 tok/s on Qwen 3.5 9B Q4_K_M. Setup requires homebrew payloads (kstuff-lite, etaHEN, ShadowMountPlus); models can be downloaded in-app from Hugging Face or sideloaded via FTP/USB.
	HN Discussion: •Project hosting on GitHub is risky due to potential Sony takedown ↑Viable for low-cost continuous background LLM tasks given power efficiency ↑Hack makes the PS5 attractive as a general-purpose computer •Similar approach could work on related cheap AMD hardware like BC250 cards ↓Limited VRAM makes it impractical and the readme oversells usability