What GPU do I need to run IFM/PAN2?
29.2B parameters, published in BF16. View on Hugging Face
PAN2 is published by IFM on Hugging Face, with 335 downloads and 161 likes to date. It's a unlisted-architecture model built for image-text-to-video, published natively in BF16.
VRAM required & cheapest live GPU fit
Required VRAM = weight size at each precision times a flat 1.2 overhead for activations and fragmentation. The KV-cache grows with context and is not in that factor; it is listed per model below where the architecture is published. Full formula and assumptions: methodology.
| Precision | Weight size | Required VRAM | Cheapest live fit | GPUs needed | Est. $/hr (full fit) |
|---|---|---|---|---|---|
| BF16 | 54.4 GB | 65.3 GB | A100 | 1 | $1.18/hr |
| cheaper alt. | RTX A5000 | 3 | $0.528/hr | ||
| FP8 (quantized) | 27.2 GB | 32.6 GB | L40 | 1 | $0.759/hr |
| cheaper alt. | RTX 4000 Ada | 2 | $0.440/hr | ||
| INT4 (quantized) | 13.6 GB | 16.3 GB | RTX A5000 | 1 | $0.176/hr |
A GPU is only matched to a row if its hardware supports that precision, and the primary recommendation is always a single-GPU fit when one exists.
INT4 requires a quantized checkpoint actually published for this model, check its Hugging Face page before relying on this row.
Cheapest way to run PAN2 at its published (BF16) precision: 1× A100, at $1.18/hr per GPU ($1.18/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
PAN2: common questions
Can PAN2 run on a single GPU?
Yes, but not on a desktop card. At BF16 it needs 65.3 GB of VRAM, more than a single 32 GB desktop card holds. The cheapest live single-card fit is a 80.0 GB A100 at $1.18/hr.
What is the least VRAM PAN2 can run in?
16.3 GB, at INT4 (quantized), which fits a 24 GB card, against 65.3 GB at BF16. That row assumes an INT4 checkpoint has actually been published for this model, so check its Hugging Face page before planning around it.
Does quantizing PAN2 lower the GPU bill?
Yes. At BF16 the cheapest live fit is one A100 at $1.18/hr. At INT4 (quantized) it drops to one RTX A5000 at $0.176/hr, provided a quantized checkpoint exists for it.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More IFM models
- K2-Horizon-MoVA-36B-A4B (37.4B, BF16)
- K2-Type-0.9B (1.1B, BF16)
Alternatives at this size
Other models for image-text-to-video within about a third of PAN2's 29.2B parameters, from other model lines.
- MiniMax-H3 (33.1B, BF16)
More on PAN2
Fits on an 80 GB GPU at BF16: every model that fits in 80 GB.
Fits on a 48 GB GPU at FP8: every model that fits in 48 GB.
Fits on a 24 GB GPU at INT4: every model that fits in 24 GB.
Related reading: A100 pricing and specs, How much VRAM you need for LLMs, and What AI inference is.
Your own GPU cloud accounts
One console for the GPU clouds you already rent from. Your keys stay on your machine.
$20 a month, flat