Hardware · Updated 17 Oct, 12:02 am IST
llama.cpp runs GGUF models natively on a jailbroken PS5
Why it matters for readers: It demonstrates you can run modern language models at home on a console and access them from your phone or laptop.
- PS5LM runs llama.cpp on the PS5's RDNA 2 GPU using Vulkan and about 11.4 GB of shared memory for models and context.1
- The console serves models over the network and accepts OpenAI /v1/chat/completions and Anthropic /v1/messages API requests.1
- It supports any architecture and quantization that llama.cpp supports, letting you load GGUF models onto the console.1
- PS5LM can also run decision-model APIs compatible with Ollaya (typed decisions with calibrated probabilities) for fast on-device answers.1
- The project targets users with a jailbroken PS5 who want a private, local model server without needing a PC, cloud service, or account.1
Get a brief like this every morning
Uzha reads hundreds of sources and gives you the stories that matter for your work, with every source linked. Free.
Get started