Hugging FaceA 111B chat model from Alibaba Qwen.
- Rating
- 2.5
- Downloads
- 5.4K
- Likes
- 4
- Parameters
- 111B
- Context
- 32K
- Licence
- other
Too big for this Mac
Its smallest size needs 30 GB of memory, more than M4 · 16 GB can give it.
Ratings
2.5 out of 5, across people's blind votes on its answers.
LM Arena ratings of October 28, 2026 (CC BY 4.0): people vote between two answers without knowing which model wrote which. The bar ends at 1488, the best any open model has. The Arena tests the full model; a small size on a Mac answers a little worse.
Qwen1.5 110B runs well from a Mac with an M3 Max and 48 GB of memory. 29 of 59 Mac configurations run it well.
Compare with
The nearest models in size.
Sizes
From bartowski/Qwen1.5-110B-Chat-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.
| Size | Download | Memory at 8K | Get |
|---|---|---|---|
| IQ1_S | 24 GB | 25 GBestimated | Open |
| IQ1_M | 26 GB | 28 GBestimated | Open |
| IQ2_XXS | 30 GB | 31 GBestimated | Open |
| IQ2_XS | 33 GB | 35 GBestimated | Open |
| IQ2_S | 34 GB | 36 GBestimated | Open |
| IQ2_M | 37 GB | 39 GBestimated | Open |
| Q2_K | 41 GB | 43 GBestimated | Open |
| IQ3_XXS | 43 GB | 45 GBestimated | Open |
| IQ3_XS | 46 GB | 48 GBestimated | Open |
| IQ3_S | 48 GB | 50 GBestimated | Open |
| Q3_K_S | 48 GB | 50 GBestimated | Open |
| IQ3_M | 50 GB | 51 GBestimated | Open |
Run It Anywhere
Qwen1.5 110B is an open model, so any app that runs GGUF files can run it. These take the IQ1_S size from bartowski straight from Hugging Face.
- LM StudioMac, Windows, Linux
Open - JanMac, Windows, Linux
Open - MstyMac, Windows, Linux
Open - OllamaMac, Windows, Linux
ollama run hf.co/bartowski/Qwen1.5-110B-Chat-GGUF:IQ1_S - llama.cppMac, Windows, Linux
llama-server -hf bartowski/Qwen1.5-110B-Chat-GGUF:IQ1_S - Docker Model RunnerMac, Windows, Linux
docker model run hf.co/bartowski/Qwen1.5-110B-Chat-GGUF:IQ1_S
On a PC Graphics Card
| Video memory | Largest size that fits | Download | Memory at 8K |
|---|---|---|---|
| 6 GB | None worth running | needs 45 GB | |
| 8 GB | None worth running | needs 45 GB | |
| 10 GB | None worth running | needs 45 GB | |
| 11 GB | None worth running | needs 45 GB | |
| 12 GB | None worth running | needs 45 GB | |
| 16 GB | None worth running | needs 45 GB | |
| 20 GB | None worth running | needs 45 GB | |
| 24 GB | None worth running | needs 45 GB | |
| 32 GB | None worth running | needs 45 GB | |
| 48 GB | IQ3_XXS | 43 GB | 45 GB |
The largest size, of those that keep three bits a weight or more, whose memory at an 8K chat is under 92% of the card's; the rest is the desktop's. Estimated from the files, the same way as for a Mac; speed on a PC is not estimated yet. A size that doesn't fit the card still runs, slower, with part of it in ordinary memory.
On every Mac
| Mac | Size | Fit | Memory | Writes |
|---|---|---|---|---|
| M6 · 16 GB | IQ1_S | Too big | 30 GB | |
| M6 · 24 GB | IQ1_S | Too big | 30 GB | |
| M6 · 32 GB | IQ1_S | Too big | 30 GB | |
| M5 · 16 GB | IQ1_S | Too big | 30 GB | |
| M5 · 24 GB | IQ1_S | Too big | 30 GB | |
| M5 · 32 GB | IQ1_S | Too big | 30 GB | |
| M5 Pro · 24 GB | IQ1_S | Too big | 30 GB | |
| M5 Pro · 48 GB | IQ2_XXS | Runs well | 31 GB | 4.3 |
| M5 Pro · 64 GB | IQ1_M | Runs great | 29 GB | 4.7 |
| M5 Max · 36 GB | IQ1_M | Tight fit | 28 GB | 7 |
| M5 Max · 48 GB | IQ2_XXS | Runs well | 31 GB | 8.4 |
| M5 Max · 64 GB | IQ1_M | Runs great | 29 GB | 9.2 |
| M5 Max · 128 GB | Q4_K_S | Runs great | 67 GB | 4.6 |
| M5 Ultra · 96 GB | Q4_K_S | Runs well | 67 GB | 9.1 |
| M5 Ultra · 256 GB | Q8_0 | Runs great | 124 GB | 6.3 |
| M5 Ultra · 512 GB | Q8_0 | Runs great | 124 GB | 6.3 |
| M4 · 16 GB | IQ1_S | Too big | 30 GB | |
| M4 · 24 GB | IQ1_S | Too big | 30 GB | |
| M4 · 32 GB | IQ1_S | Too big | 30 GB | |
| M4 Pro · 24 GB | IQ1_S | Too big | 30 GB | |
| M4 Pro · 48 GB | IQ2_XXS | Runs well | 31 GB | 3.7 |
| M4 Pro · 64 GB | IQ1_M | Runs great | 29 GB | 4.1 |
| M4 Max · 36 GB | IQ1_M | Tight fit | 28 GB | 5.9 |
| M4 Max · 48 GB | IQ2_XXS | Runs well | 31 GB | 7 |
| M4 Max · 64 GB | IQ1_M | Runs great | 29 GB | 7.6 |
| M4 Max · 96 GB | Q4_K_S | Runs well | 67 GB | 3.1 |
| M4 Max · 128 GB | Q4_K_S | Runs great | 67 GB | 4 |
| M3 · 8 GB | IQ1_S | Too big | 30 GB | |
| M3 · 16 GB | IQ1_S | Too big | 30 GB | |
| M3 · 24 GB | IQ1_S | Too big | 30 GB | |
| M3 Pro · 18 GB | IQ1_S | Too big | 30 GB | |
| M3 Pro · 36 GB | IQ1_M | Tight fit | 28 GB | 2.4 |
| M3 Max · 36 GB | IQ1_M | Tight fit | 28 GB | 4.5 |
| M3 Max · 48 GB | IQ2_XXS | Runs well | 31 GB | 5.4 |
| M3 Max · 64 GB | IQ1_M | Runs great | 29 GB | 5.9 |
| M3 Max · 96 GB | Q4_K_S | Runs well | 67 GB | 2.3 |
| M3 Max · 128 GB | Q4_K_S | Runs great | 67 GB | 3 |
| M3 Ultra · 96 GB | Q4_K_S | Runs well | 67 GB | 5.8 |
| M3 Ultra · 256 GB | Q8_0 | Runs great | 124 GB | 4 |
| M3 Ultra · 512 GB | Q8_0 | Runs great | 124 GB | 4 |
| M2 · 8 GB | IQ1_S | Too big | 30 GB | |
| M2 · 16 GB | IQ1_S | Too big | 30 GB | |
| M2 · 24 GB | IQ1_S | Too big | 30 GB | |
| M2 Pro · 16 GB | IQ1_S | Too big | 30 GB | |
| M2 Pro · 32 GB | IQ1_S | Too big | 30 GB | |
| M2 Max · 32 GB | IQ1_S | Too big | 30 GB | |
| M2 Max · 64 GB | IQ1_M | Runs great | 29 GB | 5.9 |
| M2 Max · 96 GB | Q4_K_S | Runs well | 67 GB | 3 |
| M2 Ultra · 64 GB | IQ1_M | Runs great | 29 GB | 11 |
| M2 Ultra · 128 GB | Q4_K_S | Runs great | 67 GB | 5.8 |
| M2 Ultra · 192 GB | Q8_0 | Runs great | 124 GB | 4 |
| M1 · 8 GB | IQ1_S | Too big | 30 GB | |
| M1 · 16 GB | IQ1_S | Too big | 30 GB | |
| M1 Pro · 16 GB | IQ1_S | Too big | 30 GB | |
| M1 Pro · 32 GB | IQ1_S | Too big | 30 GB | |
| M1 Max · 32 GB | IQ1_S | Too big | 30 GB | |
| M1 Max · 64 GB | IQ1_M | Runs great | 29 GB | 5.5 |
| M1 Ultra · 64 GB | IQ1_M | Runs great | 29 GB | 9.6 |
| M1 Ultra · 128 GB | Q4_K_S | Runs great | 67 GB | 5.5 |
Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.
Packaged by
2 repositories on Hugging Face carry this model for llama.cpp. Outmake reads them all as one model.
Information
- Maker
- Alibaba Qwen
- Released
- April 2024
- Parameters
- 111B
- Longest chat
- 32K tokens
- Reads pictures
- No
- Licence
- other
- Downloads this month
- 5,411
- Likes
- 4
- Sizes
- 12, 24 to 50 GB
- Kind
- Maker's release
- Source
- Qwen/Qwen1.5-110B-Chat
Badge for the model card
Shows on Qwen1.5 110B's Hugging Face card which Macs run it well, linked to this page. It updates every morning.
Files, downloads and likes read from Hugging Face on November 5, 2026 and refreshed every morning. Memory and speed are estimates until someone running Outmake measures this model on that Mac.