What GPU do I need to run HAL-X-AI/THX-02?
4.5B parameters, published in BF16. View on Hugging Face
Set up THX-02THX-02 is published by HAL-X-AI on Hugging Face, with 232 downloads and 43 likes to date. It's a THXDecisionModel model built for zero-shot-classification, 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 | 8.5 GB | 10.1 GB | RTX 4070 | 1 | $0.154/hr |
| FP8 (quantized) | 4.2 GB | 5.1 GB | RTX 4070 | 1 | $0.154/hr |
| INT4 (quantized) | 2.1 GB | 2.5 GB | RTX 3070 | 1 | $0.132/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 THX-02 at its published (BF16) precision: 1× RTX 4070, at $0.154/hr per GPU ($0.154/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
THX-02: common questions
Does THX-02 fit on a 12 GB GPU?
Yes. At BF16 it needs 10.1 GB of VRAM, so a 12 GB card holds it with 1.9 GB to spare. An 8 GB card is not enough for it at BF16.
What is the least VRAM THX-02 can run in?
2.5 GB, at INT4 (quantized), which fits a 6 GB card, against 10.1 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 THX-02 lower the GPU bill?
Yes. At BF16 the cheapest live fit is one RTX 4070 at $0.154/hr. At INT4 (quantized) it drops to one RTX 3070 at $0.132/hr, provided a quantized checkpoint exists for it.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
Alternatives at this size
Other models for zero-shot-classification within about a third of THX-02's 4.5B parameters, from other model lines.
- kanana-2-v-decide-3b-2611 (3.7B, BF16)
More on THX-02
Fits on a 12 GB GPU at BF16: every model that fits in 12 GB.
Fits on an 8 GB GPU at FP8: every model that fits in 8 GB.
Related reading: RTX 4070 pricing and specs, How much VRAM you need for LLMs, and What AI inference is.
