Memory at each precision
| Version | Weights | With context | Smallest Mac |
|---|---|---|---|
| native 4.25-bit | 65.3 GB | 70.5 GB | 96 GB |
| 3-bit lower quality | 51.2 GB | 55.3 GB | 96 GB |
| 2-bit lower quality | 36.6 GB | 39.5 GB | 64 GB |
Ships in a native low-bit format: about 65 GB of weights.
On every Apple chip
The smallest memory size of each chip that runs gpt-oss-120b, the best version that fits it, and the writing speed you can expect.
| Chip | Needs | Version | Tokens/s |
|---|---|---|---|
| M6 | no size fits | Too big | – |
| M5 Ultra | 96 GB | native 4.25-bit Runs well | 200–356 |
| M5 Max | 128 GB | native 4.25-bit Runs well | 102–182 |
| M5 Pro | 64 GB | 2-bit Runs well | 86–154 |
| M5 | no size fits | Too big | – |
| M4 Max | 128 GB | native 4.25-bit Runs well | 91–162 |
| M4 Pro | 64 GB | 2-bit Runs well | 77–137 |
| M4 | no size fits | Too big | – |
| M3 Ultra | 96 GB | native 4.25-bit Runs well | 137–243 |
| M3 Max | 96 GB | native 4.25-bit Runs well | 50–89 |
| M3 Pro | no size fits | Too big | – |
| M3 | no size fits | Too big | – |
| M2 Ultra | 128 GB | native 4.25-bit Runs well | 133–237 |
| M2 Max | 96 GB | native 4.25-bit Runs well | 67–119 |
| M2 Pro | no size fits | Too big | – |
| M2 | no size fits | Too big | – |
| M1 Ultra | 128 GB | native 4.25-bit Runs well | 133–237 |
| M1 Max | 64 GB | 2-bit Runs well | 113–200 |
| M1 Pro | no size fits | Too big | – |
| M1 | no size fits | Too big | – |
How to run it
- The easy way: download MLX Harness, free and open source. Search for gpt-oss-120b, download the version that fits your Mac, and chat. It also serves the model to your other apps over an OpenAI-compatible API.Download MLX HarnessFree · Apple Silicon · macOS 14.5 or later
- In Terminal with Apple’s MLX:
pip install mlx-lmmlx_lm.chat --model mlx-community/gpt-oss-120b-MXFP4-Q8Build: mlx-community/gpt-oss-120b-MXFP4-Q8 on Hugging Face. Other precisions sit next to it.
Step by step: download and run LLMs on a Mac · how to use MLX Harness · MLX vs Ollama vs llama.cpp
About gpt-oss-120b
| Made by | OpenAI |
|---|---|
| Parameters | 117B total · 5.1B active (mixture of experts) |
| Context | 128K tokens |
| Licence | Apache-2.0 |
| Released | 2025-08 |
| Input | Text |
| Intelligence | 11.6 on the Artificial Analysis Intelligence Index |
| Weights | openai/gpt-oss-120b |
Guide: step by step on mlx.app
Questions
- How much memory does gpt-oss-120b need?
- About 65.3 GB of weights at native 4.25-bit, plus room for the context: 70.5 GB in all. On a Mac that means 96 GB of unified memory or more, since macOS lets the GPU use about 75% of it by default.
- How fast is gpt-oss-120b on a Mac?
- Writing speed depends on memory bandwidth. The fastest Mac for it here is an M5 Ultra with 96 GB at about 200–356 tokens a second (native 4.25-bit). The smallest Mac that runs it, with 96 GB, manages about 200–356 on an M5 Ultra.
- Why does gpt-oss-120b run faster than its size suggests?
- It is a mixture-of-experts model: it holds 117B parameters in memory but reads only 5.1B for each token, so it writes at the speed of a much smaller model.
Similar models
For your model card or README
Shows the smallest Mac that runs gpt-oss-120b well and links here.
[](https://mlx.app/run/gpt-oss-120b)How it’s worked out
Memory: the model’s parameters at 8, 6, 4, 3 or 2 bits per weight plus MLX’s group scales (half a bit per weight), or its native format, with room for the context on top. By default macOS lets the GPU use about 75% of unified memory. Speed: memory bandwidth divided by the bytes read per token is the ceiling for one conversation; in llama.cpp’s public Apple Silicon benchmarks real runs reach about 45% to 80% of it, which is the range shown. These are estimates, not measurements: measure your own Mac on MLX Bench. Model figures from each official model card, chip figures from Apple’s Tech Specs. Checked 2026-10-19. Same data as mlx.app/compare.