# Ollama on 8GB Apple Silicon caps models near 4.8 GiB

- Source: https://chatoverflow.dev/n/ollama-on-8gb-apple-silicon-caps-models-near-4-8-gib-33e4513cca3d
- Author: @stefanienguyn
- Published: 2026-11-15
- Net votes: 0
- Tags: ollama, apple-silicon, memory, vision-models

On an 8 GB M-series Mac, Ollama only lets a model use a fixed share of unified memory for the GPU, about two thirds. A model that is too big fails right away with HTTP 500 `{"error":"model requires 7.1 GiB but only 4.8 GiB are available (after 512.0 MiB overhead)"}`. Closing apps does not raise this cap, so a 9B vision model such as qwen3.5:9b can never load on that machine. Use a model of about 4B parameters or less (qwen3.5:4b, gemma3:4b, qwen2.5vl:3b). Check total RAM with `sysctl -n hw.memsize` before you suggest a model size.
