What GPU do I need to run LGAI-EXAONE/K-EXAONE-236B-A23B?

237.1B parameters, published in BF16. View on Hugging Face

237.1B
Parameters
BF16
Native precision
ExaoneMoeForCausalLM
Architecture
text-generation
Pipeline

K-EXAONE-236B-A23B is published by LGAI-EXAONE on Hugging Face, with 20,951 downloads and 577 likes to date. It's a ExaoneMoeForCausalLM model built for text-generation, 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.

PrecisionWeight sizeRequired VRAMCheapest live fitGPUs neededEst. $/hr (full fit)
BF16441.6 GB530.0 GBAMD MI300X3$7.89/hr
FP8 (quantized)220.8 GB265.0 GBRTX PRO 60003$4.13/hr
INT4 (quantized)110.4 GB132.5 GBAMD MI300X1$2.63/hr
cheaper alt.RTX A50006$1.06/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 K-EXAONE-236B-A23B at its published (BF16) precision: 3× AMD MI300X, at $2.63/hr per GPU ($7.89/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.

K-EXAONE-236B-A23B: KV cache by context length

The KV cache is the memory the attention layers hold for every token of context, on top of the weights. It is not part of the flat 1.2x overhead in the table above, grows with context length and with every concurrent request, and is why a long-context deployment needs more VRAM than the table shows.

Sliding-window attention: windowed layers stop growing at the window, the full-attention layers keep growing. Cached per token: 196,608 bytes at 16-bit while the context is inside the window.

ContextKV cache, one sequence
4K0.21 GB
32K1.5 GB
128K6.0 GB
256K (model maximum)12.0 GB

Computed from the layer, head and window counts in the model's own configuration (LGAI-EXAONE/K-EXAONE-236B-A23B), at 16 bits per cached value. Each concurrent sequence needs its own cache, on top of the weights.

K-EXAONE-236B-A23B: common questions

Can K-EXAONE-236B-A23B run on a single GPU?

No. At BF16 it needs 530.0 GB of VRAM, more than a single 32 GB desktop card holds. The cheapest capable card in the live feed is a 192.0 GB AMD MI300X, and it takes 3 of them.

How many GPUs do I need to run K-EXAONE-236B-A23B?

3 at BF16. It needs 530.0 GB of VRAM and the cheapest capable live offer is a 192.0 GB AMD MI300X, so 3 of them come to $7.89/hr in total.

Does quantizing K-EXAONE-236B-A23B lower the GPU bill?

Yes. At BF16 the cheapest live fit is 3 AMD MI300X cards at $7.89/hr. At INT4 (quantized) it drops to one AMD MI300X at $2.63/hr, provided a quantized checkpoint exists for it.

Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.

LGAI-EXAONE models

Alternatives at this size

Other models for text-generation within about a third of K-EXAONE-236B-A23B's 237.1B parameters, from other model lines.

More on K-EXAONE-236B-A23B

Related reading: AMD MI300X pricing and specs, How much VRAM you need for LLMs, Serving LLMs with vLLM, vLLM vs TensorRT-LLM vs SGLang, and Best GPU for LLM inference.

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

Connect your accounts

One console for the GPU clouds you already use.

Connect your accounts for $20 a month, flat. Your provider bills the GPU.

© 2026 Aquanode. All rights reserved.

All trademarks, logos and brand names are the property of their respective owners.