Question 1

How much VRAM does Molmo-72B-0924 need?

Accepted Answer

Molmo-72B-0924 (73.3B) requires approximately 163.85678787231444 GB of VRAM at full precision (FP16), 81.9 GB at 8-bit quantization (Q8), and 41 GB at 4-bit quantization (Q4). The actual VRAM usage may be slightly higher due to KV cache and runtime overhead.

Question 2

What GPUs can run Molmo-72B-0924?

Accepted Answer

We have 5 compatible GPUs indexed for Molmo-72B-0924, including MI300X, MI350X, B200, B100, MI325X. Compatibility is determined by whether the GPU's VRAM can fit the model's memory footprint, potentially using quantization to reduce memory usage.

Question 3

What is the cheapest GPU to run Molmo-72B-0924?

Accepted Answer

The cheapest GPU to run Molmo-72B-0924 in the cloud is the MI300X at $2.39/hr. Other affordable options include MI350X ($5.49/hr) and B200 ($6.99/hr). Prices reflect the lowest on-demand rate across cloud providers.

Question 4

Can Molmo-72B-0924 run on consumer hardware?

Accepted Answer

Molmo-72B-0924 requires approximately 163.85678787231444 GB of VRAM at FP16, and even with 4-bit quantization (Q4) it needs about 41.0 GB. This exceeds the 24 GB available on the highest-end consumer GPU (RTX 4090), so Molmo-72B-0924 requires datacenter or cloud GPUs.

Question 5

What quantization options are available for Molmo-72B-0924?

Accepted Answer

Molmo-72B-0924 supports the following quantization levels: FP16 (full precision): ~163.85678787231444 GB, Q8 (8-bit): ~81.9 GB, Q6 (6-bit): ~61.4 GB, Q4 (4-bit): ~41 GB. Lower quantization reduces VRAM usage at the cost of some quality loss. Q8 offers a good balance between quality and memory savings, while Q4 provides maximum compression.

Molmo-72B-0924

VRAM Requirements

Compatible GPUs

Molmo-72B-0924 FAQ

GPU	VRAM	Best Quant	From	Rent
MI300X	192 GB	Q8	$2.39/hr	Rent
MI350X	288 GB	FP16	$5.49/hr	Rent
B200	192 GB	Q8	$6.99/hr
B100	192 GB	Q8	—
MI325X	256 GB	FP16	—

VRAM requirements for Molmo-72B-0924 at different quantization levels
Quantization	VRAM Required
FP16	163.9 GB
Q8	81.9 GB
Q6	61.4 GB
Q4	41 GB