Hugging FaceA community version of another maker's model.
- Rating
- 4.2
- Downloads
- 9.2K
- Likes
- 41
- Parameters
- 671B, 38B active
- Context
- 160K
- Licence
- mit
Too big for this Mac
Its smallest size needs 166 GB of memory, more than M4 · 16 GB can give it.
Ratings
4.2 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.
DeepSeek V3.2 runs well from a Mac with an M3 Ultra and 256 GB of memory. 4 of 59 Mac configurations run it well. It is a re-tune or merge made by someone other than the original model's maker, so its ratings, where it has any, are its own.
Compare with
The nearest models in size.
Sizes
From unsloth/DeepSeek-V3.2-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 | 161 GB | 162 GBestimated | Open |
| UD-IQ1_S | 184 GB | 185 GBestimated | Open |
| UD-IQ1_M | 199 GB | 200 GBestimated | Open |
| UD-IQ2_XXS | 217 GB | 218 GBestimated | Open |
| UD-IQ2_M | 228 GB | 229 GBestimated | Open |
| Q2_K | 245 GB | 246 GBestimated | Open |
| Q2_K_L | 246 GB | 246 GBestimated | Open |
| UD-Q2_K_XL | 247 GB | 247 GBestimated | Open |
| UD-IQ3_XXS | 273 GB | 274 GBestimated | Open |
| Q3_K_S | 290 GB | 291 GBestimated | Open |
| Q3_K_M | 320 GB | 321 GBestimated | Open |
| UD-Q3_K_XL | 321 GB | 321 GBestimated | Open |
| IQ4_XS | 358 GB | 359 GBestimated | Open |
| IQ4_NL | 379 GB | 380 GBestimated | Open |
| Q4_0 | 380 GB | 381 GBestimated | Open |
| Q4_K_S | 381 GB | 381 GBestimated | Open |
| Q4_K_M | 405 GB | 406 GBestimated | Open |
| UD-Q4_K_XL | 408 GB | 408 GBestimated | Open |
| Q4_1 | 421 GB | 421 GBestimated | Open |
| Q5_K_S | 463 GB | 463 GBestimated | Open |
| Q5_K_M | 476 GB | 477 GBestimated | Open |
| UD-Q5_K_XL | 482 GB | 482 GBestimated | Open |
| Q6_K | 551 GB | 552 GBestimated | Open |
| UD-Q6_K_XL | 574 GB | 574 GBestimated | Open |
| Q8_0 | 713 GB | 714 GBestimated | Open |
| UD-Q8_K_XL | 780 GB | 781 GBestimated | Open |
| BF16 | 1342 GB | 1343 GBestimated | Open |
Run It Anywhere
DeepSeek V3.2 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/DeepSeek-V3.2-GGUF:UD-TQ1_0 - llama.cppMac, Windows, Linux
llama-server -hf unsloth/DeepSeek-V3.2-GGUF:UD-TQ1_0 - Docker Model RunnerMac, Windows, Linux
docker model run hf.co/unsloth/DeepSeek-V3.2-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 229 GB | |
| 8 GB | None worth running | needs 229 GB | |
| 10 GB | None worth running | needs 229 GB | |
| 11 GB | None worth running | needs 229 GB | |
| 12 GB | None worth running | needs 229 GB | |
| 16 GB | None worth running | needs 229 GB | |
| 20 GB | None worth running | needs 229 GB | |
| 24 GB | None worth running | needs 229 GB | |
| 32 GB | None worth running | needs 229 GB | |
| 48 GB | None worth running | needs 229 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 | UD-TQ1_0 | Too big | 166 GB | |
| M6 · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M6 · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Pro · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Pro · 48 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Pro · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Max · 36 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Max · 48 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Max · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Max · 128 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Ultra · 96 GB | UD-TQ1_0 | Too big | 166 GB | |
| M5 Ultra · 256 GB | UD-TQ1_0 | Runs great | 166 GB | 36 |
| M5 Ultra · 512 GB | UD-Q4_K_XL | Runs well | 409 GB | 18 |
| M4 · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Pro · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Pro · 48 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Pro · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Max · 36 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Max · 48 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Max · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Max · 96 GB | UD-TQ1_0 | Too big | 166 GB | |
| M4 Max · 128 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 · 8 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Pro · 18 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Pro · 36 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Max · 36 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Max · 48 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Max · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Max · 96 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Max · 128 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Ultra · 96 GB | UD-TQ1_0 | Too big | 166 GB | |
| M3 Ultra · 256 GB | UD-TQ1_0 | Runs great | 166 GB | 20 |
| M3 Ultra · 512 GB | UD-Q4_K_XL | Runs well | 409 GB | 11 |
| M2 · 8 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 · 24 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Pro · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Pro · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Max · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Max · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Max · 96 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Ultra · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Ultra · 128 GB | UD-TQ1_0 | Too big | 166 GB | |
| M2 Ultra · 192 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 · 8 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 Pro · 16 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 Pro · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 Max · 32 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 Max · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 Ultra · 64 GB | UD-TQ1_0 | Too big | 166 GB | |
| M1 Ultra · 128 GB | UD-TQ1_0 | Too big | 166 GB |
Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.
Information
- Maker
- DeepSeek
- Released
- December 2025
- Parameters
- 671B, 38B active
- Longest chat
- 160K tokens
- Reads pictures
- No
- Licence
- mit
- Downloads this month
- 9,154
- Likes
- 41
- Sizes
- 27, 161 to 1342 GB
- Kind
- Community version
Badge for the model card
Shows on DeepSeek V3.2'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.