Outmake Models
Models

Hy Memory 30B A3B

Tencent · October 2026

Get Hugging Face

A 30B, 2.9B active chat model from Tencent.

Downloads
144K
Likes
31
Parameters
30B, 2.9B active
Context
256K
Licence
apache-2.0

Runs great on M4 · 16 GB

The TQ1_0 size from vinnipukh is the one to get: a 6.6 GB download that uses 7.7 GB of memory, and writes about 12 words a second.

Hy Memory 30B A3B runs well from a Mac with an M1 and 8 GB of memory. 59 of 59 Mac configurations run it well.

Compare with

The nearest models in size.

Sizes

From tencent/Hy-Memory-30B-A3B-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
PSIQ1_M8.0 GB8.6 GBestimatedOpen
PSIQ2_XXS9.1 GB9.7 GBestimatedOpen
Q3_K_M14 GB15 GBestimatedOpen
Q4_K_M18 GB19 GBestimatedOpen
Q6_K25 GB25 GBestimatedOpen
Q8_032 GB33 GBestimatedOpen

Run It Anywhere

Hy Memory 30B A3B is an open model, so any app that runs GGUF files can run it. These take the TQ1_0 size from vinnipukh straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/vinnipukh/Hy-Memory-30B-A3B-GGUF:TQ1_0
  • llama.cppMac, Windows, Linuxllama-server -hf vinnipukh/Hy-Memory-30B-A3B-GGUF:TQ1_0
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/vinnipukh/Hy-Memory-30B-A3B-GGUF:TQ1_0
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 15 GB
8 GBNone worth runningneeds 15 GB
10 GBNone worth runningneeds 15 GB
11 GBNone worth runningneeds 15 GB
12 GBNone worth runningneeds 15 GB
16 GBQ3_K_M14 GB15 GB
20 GBQ4_K_M18 GB19 GB
24 GBQ4_K_M18 GB19 GB
32 GBQ6_K25 GB25 GB
48 GBQ8_032 GB33 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
MacSizeFitMemoryWrites
M6 · 16 GBTQ1_0Runs great7.7 GB19
M6 · 24 GBQ3_K_MRuns well15 GB14
M6 · 32 GBQ4_K_MRuns well20 GB13
M5 · 16 GBTQ1_0Runs great7.7 GB18
M5 · 24 GBQ3_K_MRuns well15 GB13
M5 · 32 GBQ4_K_MRuns well20 GB12
M5 Pro · 24 GBQ3_K_MRuns well15 GB25
M5 Pro · 48 GBQ5_1Runs great23 GB24
M5 Pro · 64 GBQ6_KRuns great29 GB21
M5 Max · 36 GBIQ4_XSRuns great17 GB34
M5 Max · 48 GBQ5_1Runs great23 GB45
M5 Max · 64 GBQ6_KRuns great29 GB39
M5 Max · 128 GBQ8_0Runs great47 GB40
M5 Ultra · 96 GBQ8_0Runs great40 GB70
M5 Ultra · 256 GBQ8_0Runs great47 GB70
M5 Ultra · 512 GBQ8_0Runs great47 GB70
M4 · 16 GBTQ1_0Runs great7.7 GB12
M4 · 24 GBQ3_K_MRuns well15 GB9.3
M4 · 32 GBQ4_K_MRuns well20 GB8.7
M4 Pro · 24 GBQ3_K_MRuns well15 GB18
M4 Pro · 48 GBQ5_1Runs great23 GB18
M4 Pro · 64 GBQ6_KRuns great29 GB15
M4 Max · 36 GBIQ4_XSRuns great17 GB23
M4 Max · 48 GBQ5_1Runs great23 GB29
M4 Max · 64 GBQ6_KRuns great29 GB25
M4 Max · 96 GBQ8_0Runs great40 GB22
M4 Max · 128 GBQ8_0Runs great47 GB26
M3 · 8 GBQ1_0Runs well5.0 GB11
M3 · 16 GBTQ1_0Runs great7.7 GB11
M3 · 24 GBQ3_K_MRuns well15 GB8.1
M3 Pro · 18 GBPSIQ1_MRuns great8.6 GB15
M3 Pro · 36 GBIQ4_XSRuns great17 GB11
M3 Max · 36 GBIQ4_XSRuns great17 GB19
M3 Max · 48 GBQ5_1Runs great23 GB25
M3 Max · 64 GBQ6_KRuns great29 GB21
M3 Max · 96 GBQ8_0Runs great40 GB18
M3 Max · 128 GBQ8_0Runs great47 GB22
M3 Ultra · 96 GBQ8_0Runs great40 GB35
M3 Ultra · 256 GBQ8_0Runs great47 GB35
M3 Ultra · 512 GBQ8_0Runs great47 GB35
M2 · 8 GBQ1_0Runs well5.0 GB11
M2 · 16 GBTQ1_0Runs great7.7 GB11
M2 · 24 GBQ3_K_MRuns well15 GB8.1
M2 Pro · 16 GBTQ1_0Runs great7.7 GB19
M2 Pro · 32 GBQ4_K_MRuns well20 GB14
M2 Max · 32 GBQ4_K_MRuns well20 GB23
M2 Max · 64 GBQ6_KRuns great29 GB21
M2 Max · 96 GBQ8_0Runs great40 GB22
M2 Ultra · 64 GBQ6_KRuns great29 GB33
M2 Ultra · 128 GBQ8_0Runs great47 GB35
M2 Ultra · 192 GBQ8_0Runs great47 GB35
M1 · 8 GBQ1_0Runs well5.0 GB7.6
M1 · 16 GBTQ1_0Runs great7.7 GB7.2
M1 Pro · 16 GBTQ1_0Runs great7.7 GB17
M1 Pro · 32 GBQ4_K_MRuns well20 GB12
M1 Max · 32 GBQ4_K_MRuns well20 GB19
M1 Max · 64 GBQ6_KRuns great29 GB18
M1 Ultra · 64 GBQ6_KRuns great29 GB25
M1 Ultra · 128 GBQ8_0Runs great47 GB27

Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.

Packaged by

3 repositories on Hugging Face carry this model for llama.cpp. Outmake reads them all as one model.

Information

Maker
Tencent
Released
October 2026
Parameters
30B, 2.9B active
Longest chat
256K tokens
Reads pictures
No
Licence
apache-2.0
Downloads this month
143,853
Likes
31
Sizes
6, 8.0 to 32 GB
Kind
Maker's release

Badge for the model card

Shows on Hy Memory 30B A3B's Hugging Face card which Macs run it well, linked to this page. It updates every morning.

Runs on Mac badge

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.

Your Computer

Every model is judged for this Mac. Both are under About This Mac, in the Apple menu.

Every Mac · Every graphics card

Civitai

Open in OutmakeHugging Face

New to Outmake? Download the free beta