Gemma 4 26B (4B active)

on m5 pro (18 core cpu, 20 core gpu) · Ollama

May 1, 2026

Throughput

68.8 t/s gen · 204.1 tokens/s prompt processing (prefill)Prompt processing measures input prefill speed; it does not tell you the time to first token.

System RAM

48 GB

Summary

User reports 204.07 t/s prompt eval and 68.76 t/s generation on an M5 Pro with 48 GB RAM.