My Projects

Local AI on my PS5: one Gemma 4 run at 61.1 tokens/s

I tried three local language models through PS5LM Plus. One Gemma 4 run reached about 61.1 tokens per second; I have not yet published matched timings for the other models.