Outmake Models
Models

Llama 2 70B

TheBloke · September 2023

Get Hugging Face

A 69B chat model from TheBloke.

Rating
2.1
Downloads
18K
Likes
151
Parameters
69B
Context
4K
Licence
llama2

Too big for this Mac

Its smallest size needs 15 GB of memory, more than M4 · 16 GB can give it.

Ratings

2.1 out of 5, across people's blind votes on its answers.

Text
1171
Coding
1179

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.

Llama 2 70B runs well from a Mac with an M2 and 24 GB of memory. 47 of 59 Mac configurations run it well.

Compare with

The nearest models in size.

Sizes

From mradermacher/Llama-2-70b-chat-hf-i1-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
IQ1_S15 GB16 GBestimatedOpen
IQ1_M16 GB18 GBestimatedOpen
IQ2_XXS18 GB20 GBestimatedOpen
IQ2_XS20 GB22 GBestimatedOpen
IQ2_S21 GB23 GBestimatedOpen
IQ2_M23 GB25 GBestimatedOpen
Q2_K25 GB27 GBestimatedOpen
IQ3_XXS27 GB28 GBestimatedOpen
IQ3_XS28 GB30 GBestimatedOpen
IQ3_S30 GB32 GBestimatedOpen
IQ3_M31 GB33 GBestimatedOpen
IQ4_XS37 GB38 GBestimatedOpen

Run It Anywhere

Llama 2 70B is an open model, so any app that runs GGUF files can run it. These take the IQ1_S size from mradermacher straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/mradermacher/Llama-2-70b-chat-hf-i1-GGUF:IQ1_S
  • llama.cppMac, Windows, Linuxllama-server -hf mradermacher/Llama-2-70b-chat-hf-i1-GGUF:IQ1_S
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/mradermacher/Llama-2-70b-chat-hf-i1-GGUF:IQ1_S
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 28 GB
8 GBNone worth runningneeds 28 GB
10 GBNone worth runningneeds 28 GB
11 GBNone worth runningneeds 28 GB
12 GBNone worth runningneeds 28 GB
16 GBNone worth runningneeds 28 GB
20 GBNone worth runningneeds 28 GB
24 GBNone worth runningneeds 28 GB
32 GBIQ3_S30 GB32 GB
48 GBIQ4_XS37 GB38 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 GBIQ1_SToo big15 GB
M6 · 24 GBIQ1_SRuns well15 GB3.8
M6 · 32 GBIQ1_SRuns great15 GB3.8
M5 · 16 GBIQ1_SToo big15 GB
M5 · 24 GBIQ1_SRuns well15 GB3.5
M5 · 32 GBIQ1_SRuns great15 GB3.5
M5 Pro · 24 GBIQ1_SRuns well15 GB6.9
M5 Pro · 48 GBIQ3_MRuns well32 GB4.2
M5 Pro · 64 GBQ4_K_MRuns well42 GB3.3
M5 Max · 36 GBIQ1_MRuns great17 GB9.6
M5 Max · 48 GBIQ3_MRuns well32 GB8.2
M5 Max · 64 GBQ4_K_MRuns well42 GB6.5
M5 Max · 128 GBQ5_K_MRuns great50 GB5.8
M5 Ultra · 96 GBQ4_K_MRuns great42 GB13
M5 Ultra · 256 GBQ5_K_MRuns great50 GB11
M5 Ultra · 512 GBQ5_K_MRuns great50 GB11
M4 · 16 GBIQ1_SToo big15 GB
M4 · 24 GBIQ1_SRuns well15 GB2.7
M4 · 32 GBIQ1_SRuns great15 GB2.7
M4 Pro · 24 GBIQ1_SRuns well15 GB5.8
M4 Pro · 48 GBIQ3_MRuns well32 GB3.6
M4 Pro · 64 GBQ4_K_MRuns well42 GB2.9
M4 Max · 36 GBIQ1_MRuns great17 GB7.8
M4 Max · 48 GBIQ3_MRuns well32 GB6.8
M4 Max · 64 GBQ4_K_MRuns well42 GB5.6
M4 Max · 96 GBQ4_K_MRuns great42 GB4.3
M4 Max · 128 GBQ5_K_MRuns great50 GB5
M3 · 8 GBIQ1_SToo big15 GB
M3 · 16 GBIQ1_SToo big15 GB
M3 · 24 GBIQ1_SRuns well15 GB2.3
M3 Pro · 18 GBIQ1_SToo big15 GB
M3 Pro · 36 GBIQ1_MRuns great17 GB3.2
M3 Max · 36 GBIQ1_MRuns great17 GB6.1
M3 Max · 48 GBIQ3_MRuns well32 GB5.2
M3 Max · 64 GBQ4_K_MRuns well42 GB4.3
M3 Max · 96 GBQ4_K_MRuns great42 GB3.2
M3 Max · 128 GBQ5_K_MRuns great50 GB3.8
M3 Ultra · 96 GBQ4_K_MRuns great42 GB8
M3 Ultra · 256 GBQ5_K_MRuns great50 GB7.1
M3 Ultra · 512 GBQ5_K_MRuns great50 GB7.1
M2 · 8 GBIQ1_SToo big15 GB
M2 · 16 GBIQ1_SToo big15 GB
M2 · 24 GBIQ1_SRuns well15 GB2.3
M2 Pro · 16 GBIQ1_SToo big15 GB
M2 Pro · 32 GBIQ1_SRuns great15 GB4.4
M2 Max · 32 GBIQ1_SRuns great15 GB8.2
M2 Max · 64 GBQ4_K_MRuns well42 GB4.3
M2 Max · 96 GBQ4_K_MRuns great42 GB4.3
M2 Ultra · 64 GBQ4_K_MRuns well42 GB8
M2 Ultra · 128 GBQ5_K_MRuns great50 GB7.1
M2 Ultra · 192 GBQ5_K_MRuns great50 GB7.1
M1 · 8 GBIQ1_SToo big15 GB
M1 · 16 GBIQ1_SToo big15 GB
M1 Pro · 16 GBIQ1_SToo big15 GB
M1 Pro · 32 GBIQ1_SRuns great15 GB4.2
M1 Max · 32 GBIQ1_SRuns great15 GB7.6
M1 Max · 64 GBQ4_K_MRuns well42 GB4.1
M1 Ultra · 64 GBQ4_K_MRuns well42 GB7.3
M1 Ultra · 128 GBQ5_K_MRuns great50 GB6.6

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

Packaged by

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

Information

Maker
TheBloke
Released
September 2023
Parameters
69B
Longest chat
4K tokens
Reads pictures
No
Licence
llama2
Downloads this month
17,790
Likes
151
Sizes
12, 15 to 37 GB
Kind
Maker's release

Badge for the model card

Shows on Llama 2 70B'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