Memory at each precision
| Version | Weights | With context | Smallest Mac |
|---|---|---|---|
| 8-bit | 313 GB | 339 GB | 512 GB |
| 6-bit | 240 GB | 259 GB | 512 GB |
| 4-bit | 166 GB | 179 GB | 256 GB |
| 3-bit lower quality | 129 GB | 139 GB | 192 GB |
| 2-bit lower quality | 92.2 GB | 99.6 GB | 192 GB |
On every Apple chip
The smallest memory size of each chip that runs Hy3, 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 | 256 GB | 4-bit Runs well | 46–81 |
| M5 Max | 128 GB | 2-bit Only just | 42–74 |
| M5 Pro | no size fits | Too big | – |
| M5 | no size fits | Too big | – |
| M4 Max | 128 GB | 2-bit Only just | 37–66 |
| M4 Pro | no size fits | Too big | – |
| M4 | no size fits | Too big | – |
| M3 Ultra | 256 GB | 4-bit Runs well | 31–56 |
| M3 Max | 128 GB | 2-bit Only just | 27–48 |
| M3 Pro | no size fits | Too big | – |
| M3 | no size fits | Too big | – |
| M2 Ultra | 192 GB | 3-bit Runs well | 39–70 |
| M2 Max | no size fits | Too big | – |
| M2 Pro | no size fits | Too big | – |
| M2 | no size fits | Too big | – |
| M1 Ultra | 128 GB | 2-bit Only just | 55–97 |
| M1 Max | no size fits | Too big | – |
| 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 Hy3, 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/Hy3-OptiQ-2bitBuild: mlx-community/Hy3-OptiQ-2bit 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 Hy3
| Made by | Tencent |
|---|---|
| Parameters | 295B total · 21B active (mixture of experts) |
| Context | 256K tokens |
| Licence | Apache-2.0 |
| Released | 2026-07 |
| Input | Text |
| Intelligence | 25.3 on the Artificial Analysis Intelligence Index |
| Weights | tencent/Hy3 |
Questions
- How much memory does Hy3 need?
- About 166 GB of weights at 4-bit, plus room for the context: 179 GB in all. On a Mac that means 256 GB of unified memory or more, since macOS lets the GPU use about 75% of it by default.
- How fast is Hy3 on a Mac?
- Writing speed depends on memory bandwidth. The fastest Mac for it here is an M5 Ultra with 256 GB at about 46–81 tokens a second (4-bit). The smallest Mac that runs it, with 256 GB, manages about 46–81 on an M5 Ultra.
- Why does Hy3 run faster than its size suggests?
- It is a mixture-of-experts model: it holds 295B parameters in memory but reads only 21B 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 Hy3 well and links here.
[](https://mlx.app/run/hy3)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.