PS5LM · 2026-10-16 · notable
PS5LM v0.9.0 — llama.cpp runs on a jailbroken PS5's GPU
PS5LM v0.9.0 runs llama.cpp on a jailbroken PS5's GPU through Vulkan and serves GGUF models to your network over OpenAI- and Anthropic-style APIs. Qwen 3.8 27B runs at about 20 tokens/s on a PS5 Slim.
A homebrew app turns a jailbroken PS5 into a local LLM server for your home network.
Key specs
| Qwen 3.8 27 b | ~20 tok/s |
|---|---|
| Qwen 3.5 9 b | 55 tok/s |
What is it?
PS5LM v0.9.0 moves llama.cpp onto the PS5 GPU as its own console app, with a new TV interface of Home, Models, API, Console and Settings screens. The first release on October 2 ran small Qwen models on the CPU only, with a 1.6 GB model cap.
How does it work?
The app uses llama.cpp's Vulkan backend on the console's RDNA 2 graphics, with 11.4 GB of shared memory for the model and its context. Other devices call it on port 8081 through /v1/chat/completions, /v1/messages or a browser chat page, and Ollaya-style decision models answer on /v1/systemone.
Why does it matter?
Owners of a jailbroken PS5 get a private, always-on model server with no PC, cloud or account while it runs, and the project's tests put a 27B model at about 20 tokens/s on the console. It only works on firmware 7.00 to 13.60, and v0.9.0 has a known issue where a decision model can stop after about 55 requests.
Who is it for?
local-LLM tinkerers with a jailbroken PS5
Try it
Copy the PPSA99575 folder from ps5lm-app-v0.9.0.zip to /data/homebrew, then open http://<console>:8081