Question 1

What AI models can run on M4 (24 GB)?

Accepted Answer

The M4 (24 GB) (18 GB VRAM) can run models including Llama-2-7b-chat, HARC-Llama-3.1-8B-Instruct, FastContext-1.0-4B-RL, Llama-3.2-1B-Instruct-QLORA_INT4_EO8, Llama-2-7b and more — 10 compatible models indexed so far. Compatible models are determined by whether the model's memory footprint fits within the GPU's 18 GB VRAM, potentially using quantization (Q8, Q6, or Q4) to reduce memory usage.

Question 2

What is the cheapest cloud option for M4 (24 GB)?

Accepted Answer

We don't currently have cloud pricing data for the M4 (24 GB). This GPU may not be widely available from major cloud providers, or pricing data hasn't been synced yet.

Question 3

Is M4 (24 GB) good for fine-tuning?

Accepted Answer

The M4 (24 GB) can handle fine-tuning with some constraints. With 18 GB of VRAM, it's suitable for LoRA/QLoRA fine-tuning of models up to ~7B parameters. Full fine-tuning is limited to smaller models (~1-3B parameters). Techniques like gradient checkpointing and mixed precision help maximize what fits.

Question 4

How much VRAM does M4 (24 GB) have?

Accepted Answer

The M4 (24 GB) has 18 GB of VRAM (LPDDR5X). It also offers 120 GB/s memory bandwidth, 8.4 TFLOPS of FP16 compute.

Question 5

What quantization formats work on M4 (24 GB)?

Accepted Answer

The M4 (24 GB) supports the following quantization formats: FP16 (full precision), Q8 (8-bit), Q6 (6-bit), Q4 (4-bit). FP16 works for models up to ~3B parameters. Q4 quantization can fit models requiring up to ~72 GB at FP16 into 18 GB.

M4 (24 GB)

Specifications

Cloud pricing

Compatible models

Compare this GPU

Frequently asked questions

VRAM	18 GB
Memory type	LPDDR5X
Bandwidth	120 GB/s
FP16 TFLOPS	8.4
FP8 support	No