Hugging FaceA community version of another maker's model.
- Rating
- 4.4
- Downloads
- 26K
- Likes
- 289
- Parameters
- 284B, 14B active
- Context
- 1M
- Licence
- mit
Too big for this Mac
Its smallest size needs 68 GB of memory, more than M4 · 16 GB can give it.
Ratings
4.4 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 V4 Flash runs well from a Mac with an M2 Max and 96 GB of memory. 15 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-V4-Flash-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.
| Size | Download | Memory at 8K | Get |
|---|---|---|---|
| UD-IQ1_S | 83 GB | 83 GBestimated | Open |
| UD-IQ1_M | 87 GB | 87 GBestimated | Open |
| UD-IQ2_XXS | 91 GB | 91 GBestimated | Open |
| UD-IQ2_M | 91 GB | 91 GBestimated | Open |
| UD-Q2_K_XL | 97 GB | 97 GBestimated | Open |
| UD-IQ3_XXS | 103 GB | 103 GBestimated | Open |
| UD-IQ3_S | 117 GB | 117 GBestimated | Open |
| UD-Q3_K_M | 129 GB | 129 GBestimated | Open |
| UD-Q3_K_XL | 129 GB | 130 GBestimated | Open |
| UD-IQ4_NL | 138 GB | 138 GBestimated | Open |
| UD-IQ4_XS | 138 GB | 138 GBestimated | Open |
| UD-Q4_K_XL | 155 GB | 155 GBestimated | Open |
Run It Anywhere
DeepSeek V4 Flash is an open model, so any app that runs GGUF files can run it. These take the UD-IQ1_S 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-V4-Flash-GGUF:UD-IQ1_S - llama.cppMac, Windows, Linux
llama-server -hf unsloth/DeepSeek-V4-Flash-GGUF:UD-IQ1_S - Docker Model RunnerMac, Windows, Linux
docker model run hf.co/unsloth/DeepSeek-V4-Flash-GGUF:UD-IQ1_S
On a PC Graphics Card
| Video memory | Largest size that fits | Download | Memory at 8K |
|---|---|---|---|
| 6 GB | None worth running | needs 91 GB | |
| 8 GB | None worth running | needs 91 GB | |
| 10 GB | None worth running | needs 91 GB | |
| 11 GB | None worth running | needs 91 GB | |
| 12 GB | None worth running | needs 91 GB | |
| 16 GB | None worth running | needs 91 GB | |
| 20 GB | None worth running | needs 91 GB | |
| 24 GB | None worth running | needs 91 GB | |
| 32 GB | None worth running | needs 91 GB | |
| 48 GB | None worth running | needs 91 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 | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M6 · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M6 · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Pro · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Pro · 48 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Pro · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Max · 36 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Max · 48 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Max · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M5 Max · 128 GB | IQ1_M | Runs great | 68 GB | 56 |
| M5 Ultra · 96 GB | IQ1_M | Runs well | 68 GB | 92 |
| M5 Ultra · 256 GB | UD-Q4_K_XL | Runs great | 162 GB | 50 |
| M5 Ultra · 512 GB | Q8_0 | Runs great | 309 GB | 35 |
| M4 · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Pro · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Pro · 48 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Pro · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Max · 36 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Max · 48 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Max · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M4 Max · 96 GB | IQ1_M | Runs well | 68 GB | 29 |
| M4 Max · 128 GB | IQ1_M | Runs great | 68 GB | 33 |
| M3 · 8 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 Pro · 18 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 Pro · 36 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 Max · 36 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 Max · 48 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 Max · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M3 Max · 96 GB | IQ1_M | Runs well | 68 GB | 25 |
| M3 Max · 128 GB | IQ1_M | Runs great | 68 GB | 30 |
| M3 Ultra · 96 GB | IQ1_M | Runs well | 68 GB | 44 |
| M3 Ultra · 256 GB | UD-Q4_K_XL | Runs great | 162 GB | 28 |
| M3 Ultra · 512 GB | Q8_0 | Runs great | 309 GB | 20 |
| M2 · 8 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 · 24 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 Pro · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 Pro · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 Max · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 Max · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 Max · 96 GB | IQ1_M | Runs well | 68 GB | 30 |
| M2 Ultra · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M2 Ultra · 128 GB | IQ1_M | Runs great | 68 GB | 43 |
| M2 Ultra · 192 GB | UD-Q3_K_M | Runs great | 131 GB | 31 |
| M1 · 8 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 Pro · 16 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 Pro · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 Max · 32 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 Max · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 Ultra · 64 GB | DeepSeek-V4-Flash-IQ1_S-XL | Too big | 68 GB | |
| M1 Ultra · 128 GB | IQ1_M | Runs great | 68 GB | 32 |
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
- DeepSeek
- Released
- April 2026
- Parameters
- 284B, 14B active
- Longest chat
- 1M tokens
- Reads pictures
- No
- Licence
- mit
- Downloads this month
- 25,963
- Likes
- 289
- Sizes
- 12, 83 to 155 GB
- Kind
- Community version
Badge for the model card
Shows on DeepSeek V4 Flash'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.