What GPU do I need to run LGAI-EXAONE/EXAONE-3.5-32B-Instruct?
A 32B language model for chat and instruction-following. 32.0B parameters, published in F32. View on Hugging Face
EXAONE-3.5-32B-Instruct is published by LGAI-EXAONE on Hugging Face, with 36,125 downloads and 130 likes to date. It's a ExaoneForCausalLM model built for text-generation, published natively in F32.
What EXAONE-3.5-32B-Instruct is
EXAONE-3.5-32B-Instruct is a 32B-parameter language model published by LG AI Research on Hugging Face. It is released under Custom license.
License note: a lab-specific license (tagged "other" on Hugging Face); read the model card's own license section before commercial use. Facts in this section are sourced from EXAONE-3.5-32B-Instruct's Hugging Face model card, not benchmarked by Aquanode.
What it's used for
- Chat assistants
- Instruction following
- Synthetic data generation
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) |
|---|---|---|---|---|---|
| FP32 | 119.2 GB | 143.1 GB | AMD MI300X | 1 | $2.63/hr |
| cheaper alt. | RTX A5000 | 6 | $1.06/hr | ||
| FP8 (quantized) | 29.8 GB | 35.8 GB | L40 | 1 | $0.759/hr |
| cheaper alt. | RTX 4000 Ada | 2 | $0.440/hr | ||
| INT4 (quantized) | 14.9 GB | 17.9 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 EXAONE-3.5-32B-Instruct at its published (F32) precision: 1× AMD MI300X, at $2.63/hr per GPU ($2.63/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
EXAONE-3.5-32B-Instruct: 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.
Grouped-query attention: every layer caches keys and values for a few shared KV heads. Cached per token: 262,144 bytes at 16-bit.
| Context | KV cache, one sequence |
|---|---|
| 4K | 1.0 GB |
| 32K | 8.0 GB |
Computed from the layer, head and window counts in the model's own configuration (LGAI-EXAONE/EXAONE-3.5-32B-Instruct), at 16 bits per cached value. Each concurrent sequence needs its own cache, on top of the weights.
EXAONE-3.5-32B-Instruct: common questions
Can EXAONE-3.5-32B-Instruct run on a single GPU?
Yes, but not on a desktop card. At FP32 it needs 143.1 GB of VRAM, more than a single 32 GB desktop card holds. The cheapest live single-card fit is a 192.0 GB AMD MI300X at $2.63/hr.
Can EXAONE-3.5-32B-Instruct run in 16-bit instead of FP32?
Yes. Its published weights are FP32, 119.2 GB, or 143.1 GB once inference overhead is added. Loading the same weights in 16 bits halves that to 59.6 GB, or 71.5 GB with overhead. How much accuracy the cast costs is model-specific and is not measured here.
What is the least VRAM EXAONE-3.5-32B-Instruct can run in?
17.9 GB, at INT4 (quantized), which fits a 24 GB card, against 143.1 GB at FP32. 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 EXAONE-3.5-32B-Instruct lower the GPU bill?
Yes. At FP32 the cheapest live fit is one AMD MI300X at $2.63/hr. At INT4 (quantized) it drops to one RTX A5000 at $0.176/hr, provided a quantized checkpoint exists for it.
How to run EXAONE-3.5-32B-Instruct
Run EXAONE-3.5-32B-Instruct with vLLM
Generic example, not from the model's own docs: adjust flags (quantization, context length, parallelism) for your setup.
vllm serve LGAI-EXAONE/EXAONE-3.5-32B-Instruct --tensor-parallel-size 1Run EXAONE-3.5-32B-Instruct with Ollama
Verified against Ollama's own library listing.
ollama run exaone3.5:32bRun EXAONE-3.5-32B-Instruct with GGUF quantizations
Prebuilt GGUF weights published at bartowski/EXAONE-3.5-32B-Instruct-GGUF. Run with llama.cpp's llama-server or load the repo directly in LM Studio.
llama-server -hf bartowski/EXAONE-3.5-32B-Instruct-GGUFSource: https://huggingface.co/bartowski/EXAONE-3.5-32B-Instruct-GGUF
Deploy EXAONE-3.5-32B-Instruct on Aquanode
Aquanode has no one-click deploy template for EXAONE-3.5-32B-Instruct; you install the inference engine yourself with the commands below. Aquanode sells GPU pods billed per second, not a hosted inference API.
- Launch a bare GPU pod sized to the requirement above (1× AMD MI300X or larger).
- Open a terminal on the pod, or save one of the commands above as a startup script so it runs automatically the first time the pod boots.
- Run the command and connect to the resulting endpoint.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More EXAONE 3.5 models
- EXAONE-3.5-7.8B-Instruct (7.8B, F32)
- EXAONE-3.0-7.8B-Instruct (7.8B, F32)
- EXAONE-3.5-2.4B-Instruct (2.4B, F32)
Alternatives at this size
Other models for text-generation within about a third of EXAONE-3.5-32B-Instruct's 32.0B parameters, from other model lines.
- Qwen3-32B (32.8B, BF16)
- GLM-4.7-Flash (31.2B, BF16)
- NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 (31.6B, BF16)
- granite-4.1-30b (28.9B, BF16)
- Ornith-1.5-35B-A3B (36.0B, BF16)
More on EXAONE-3.5-32B-Instruct
Fits on a 192 GB GPU at FP32: every model that fits in 192 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.
Best chat and assistants models: how EXAONE-3.5-32B-Instruct ranks against the rest.
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