Hugging FaceA 80B, 3.9B active chat model from Alibaba Qwen.
- Rating
- 4.0
- Downloads
- 133K
- Likes
- 312
- Parameters
- 80B, 3.9B active
- Context
- 256K
- Licence
- apache-2.0
Too big for this Mac
Its smallest size needs 20 GB of memory, more than M4 · 16 GB can give it.
Ratings
4.0 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.
Qwen3 Next 80B A3B runs well from a Mac with an M1 Pro and 32 GB of memory. 40 of 59 Mac configurations run it well.
Compare with
The nearest models in size.
Sizes
From unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.
| Size | Download | Memory at 8K | Get |
|---|---|---|---|
| UD-TQ1_0 | 20 GB | 21 GBestimated | Open |
| UD-IQ1_S | 23 GB | 23 GBestimated | Open |
| UD-IQ1_M | 24 GB | 25 GBestimated | Open |
| UD-IQ2_XXS | 26 GB | 26 GBestimated | Open |
| Q2_K | 29 GB | 29 GBestimated | Open |
| Q2_K_L | 29 GB | 29 GBestimated | Open |
| UD-Q2_K_XL | 30 GB | 30 GBestimated | Open |
| UD-IQ3_XXS | 33 GB | 33 GBestimated | Open |
| Q3_K_S | 35 GB | 35 GBestimated | Open |
| UD-Q3_K_XL | 36 GB | 36 GBestimated | Open |
| Q3_K_M | 38 GB | 39 GBestimated | Open |
| IQ4_XS | 43 GB | 43 GBestimated | Open |
| IQ4_NL | 45 GB | 45 GBestimated | Open |
| Q4_0 | 45 GB | 46 GBestimated | Open |
| Q4_K_S | 46 GB | 46 GBestimated | Open |
| UD-Q4_K_XL | 46 GB | 46 GBestimated | Open |
| Q4_1 | 50 GB | 50 GBestimated | Open |
| Q5_K_S | 55 GB | 55 GBestimated | Open |
| UD-Q5_K_XL | 57 GB | 57 GBestimated | Open |
| UD-Q6_K_XL | 69 GB | 69 GBestimated | Open |
| UD-Q8_K_XL | 93 GB | 93 GBestimated | Open |
| BF16 | 159 GB | 160 GBestimated | Open |
Run It Anywhere
Qwen3 Next 80B A3B is an open model, so any app that runs GGUF files can run it. These take the UD-TQ1_0 size from unsloth straight from Hugging Face.
- LM StudioMac, Windows, Linux
Open - JanMac, Windows, Linux
Open - MstyMac, Windows, Linux
Open - OllamaMac, Windows, Linux
ollama run hf.co/unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF:UD-TQ1_0 - llama.cppMac, Windows, Linux
llama-server -hf unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF:UD-TQ1_0 - Docker Model RunnerMac, Windows, Linux
docker model run hf.co/unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF:UD-TQ1_0
On a PC Graphics Card
| Video memory | Largest size that fits | Download | Memory at 8K |
|---|---|---|---|
| 6 GB | None worth running | needs 30 GB | |
| 8 GB | None worth running | needs 30 GB | |
| 10 GB | None worth running | needs 30 GB | |
| 11 GB | None worth running | needs 30 GB | |
| 12 GB | None worth running | needs 30 GB | |
| 16 GB | None worth running | needs 30 GB | |
| 20 GB | None worth running | needs 30 GB | |
| 24 GB | None worth running | needs 30 GB | |
| 32 GB | UD-Q2_K_XL | 30 GB | 30 GB |
| 48 GB | UD-Q4_K_XL | 46 GB | 46 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 | 20 GB | |
| M6 · 24 GB | IQ1_M | Tight fit | 18 GB | 34 |
| M6 · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 32 |
| M5 · 16 GB | IQ1_S | Too big | 20 GB | |
| M5 · 24 GB | IQ1_M | Tight fit | 18 GB | 33 |
| M5 · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 30 |
| M5 Pro · 24 GB | IQ1_M | Tight fit | 18 GB | 58 |
| M5 Pro · 48 GB | UD-Q2_K_XL | Runs well | 31 GB | 45 |
| M5 Pro · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 45 |
| M5 Max · 36 GB | IQ1_S | Runs great | 17 GB | 79 |
| M5 Max · 48 GB | UD-Q2_K_XL | Runs well | 31 GB | 76 |
| M5 Max · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 76 |
| M5 Max · 128 GB | Q6_K | Runs great | 67 GB | 51 |
| M5 Ultra · 96 GB | UD-Q4_K_XL | Runs great | 46 GB | 97 |
| M5 Ultra · 256 GB | Q8_0 | Runs great | 89 GB | 83 |
| M5 Ultra · 512 GB | Q8_0 | Runs great | 89 GB | 83 |
| M4 · 16 GB | IQ1_S | Too big | 20 GB | |
| M4 · 24 GB | IQ1_M | Tight fit | 18 GB | 20 |
| M4 · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 19 |
| M4 Pro · 24 GB | IQ1_M | Tight fit | 18 GB | 33 |
| M4 Pro · 48 GB | UD-Q2_K_XL | Runs well | 31 GB | 28 |
| M4 Pro · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 28 |
| M4 Max · 36 GB | IQ1_S | Runs great | 17 GB | 39 |
| M4 Max · 48 GB | UD-Q2_K_XL | Runs well | 31 GB | 38 |
| M4 Max · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 38 |
| M4 Max · 96 GB | UD-Q4_K_XL | Runs great | 46 GB | 29 |
| M4 Max · 128 GB | Q6_K | Runs great | 67 GB | 30 |
| M3 · 8 GB | IQ1_S | Too big | 20 GB | |
| M3 · 16 GB | IQ1_S | Too big | 20 GB | |
| M3 · 24 GB | IQ1_M | Tight fit | 18 GB | 19 |
| M3 Pro · 18 GB | IQ1_S | Too big | 20 GB | |
| M3 Pro · 36 GB | IQ1_S | Runs great | 17 GB | 25 |
| M3 Max · 36 GB | IQ1_S | Runs great | 17 GB | 38 |
| M3 Max · 48 GB | UD-Q2_K_XL | Runs well | 31 GB | 37 |
| M3 Max · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 37 |
| M3 Max · 96 GB | UD-Q4_K_XL | Runs great | 46 GB | 26 |
| M3 Max · 128 GB | Q6_K | Runs great | 67 GB | 27 |
| M3 Ultra · 96 GB | UD-Q4_K_XL | Runs great | 46 GB | 44 |
| M3 Ultra · 256 GB | Q8_0 | Runs great | 89 GB | 40 |
| M3 Ultra · 512 GB | Q8_0 | Runs great | 89 GB | 40 |
| M2 · 8 GB | IQ1_S | Too big | 20 GB | |
| M2 · 16 GB | IQ1_S | Too big | 20 GB | |
| M2 · 24 GB | IQ1_M | Tight fit | 18 GB | 19 |
| M2 Pro · 16 GB | IQ1_S | Too big | 20 GB | |
| M2 Pro · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 28 |
| M2 Max · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 40 |
| M2 Max · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 36 |
| M2 Max · 96 GB | UD-Q4_K_XL | Runs great | 46 GB | 31 |
| M2 Ultra · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 48 |
| M2 Ultra · 128 GB | Q6_K | Runs great | 67 GB | 39 |
| M2 Ultra · 192 GB | Q8_0 | Runs great | 89 GB | 40 |
| M1 · 8 GB | IQ1_S | Too big | 20 GB | |
| M1 · 16 GB | IQ1_S | Too big | 20 GB | |
| M1 Pro · 16 GB | IQ1_S | Too big | 20 GB | |
| M1 Pro · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 22 |
| M1 Max · 32 GB | UD-TQ1_0 | Runs well | 21 GB | 30 |
| M1 Max · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 27 |
| M1 Ultra · 64 GB | UD-Q2_K_XL | Runs great | 31 GB | 33 |
| M1 Ultra · 128 GB | Q6_K | Runs great | 67 GB | 29 |
Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.
Packaged by
5 repositories on Hugging Face carry this model for llama.cpp. Outmake reads them all as one model.
Information
- Maker
- Alibaba Qwen
- Released
- September 2025
- Parameters
- 80B, 3.9B active
- Longest chat
- 256K tokens
- Reads pictures
- No
- Licence
- apache-2.0
- Downloads this month
- 133,396
- Likes
- 312
- Sizes
- 22, 20 to 159 GB
- Kind
- Maker's release
Badge for the model card
Shows on Qwen3 Next 80B A3B'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.