What GPU do I need to run HuggingFaceTB/SmolLM2-360M-Instruct?
A 362M language model for chat and instruction-following. 362M parameters, published in BF16. View on Hugging Face
SmolLM2-360M-Instruct is published by HuggingFaceTB on Hugging Face, with 307,619 downloads and 216 likes to date. It's a LlamaForCausalLM model built for text-generation, published natively in BF16.
What SmolLM2-360M-Instruct is
SmolLM2-360M-Instruct is a 362M-parameter language model published by Hugging Face on Hugging Face. It is released under Apache 2.0.
License note: fully permissive, no gating and no usage restrictions. Facts in this section are sourced from SmolLM2-360M-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) |
|---|---|---|---|---|---|
| BF16 | 0.7 GB | 0.8 GB | RTX 3070 | 1 | $0.088/hr |
| FP8 (quantized) | 0.3 GB | 0.4 GB | RTX 4070 | 1 | $0.154/hr |
| INT4 (quantized) | 0.2 GB | 0.2 GB | RTX 3070 | 1 | $0.088/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 SmolLM2-360M-Instruct at its published (BF16) precision: 1× RTX 3070, at $0.088/hr per GPU ($0.088/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
SmolLM2-360M-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: 40,960 bytes at 16-bit.
| Context | KV cache, one sequence |
|---|---|
| 4K | 0.16 GB |
| 8K (model maximum) | 0.31 GB |
Computed from the layer, head and window counts in the model's own configuration (HuggingFaceTB/SmolLM2-360M-Instruct), at 16 bits per cached value. Each concurrent sequence needs its own cache, on top of the weights.
SmolLM2-360M-Instruct: common questions
How much VRAM does SmolLM2-360M-Instruct need?
0.8 GB at BF16, 0.4 GB at FP8 (quantized), 0.2 GB at INT4 (quantized). All of those sit under 6 GB, the smallest capacity this page reasons about, so VRAM is not what limits where this model runs. The 0.7 GB of weights plus inference overhead is the whole requirement.
How many copies of SmolLM2-360M-Instruct fit on one RTX 3070?
9, by VRAM alone. That card carries 8.0 GB and one copy needs 0.8 GB at BF16, on a live rate of $0.088/hr for the whole card. Throughput is not modelled here, so 9 copies is not 9 times the requests served.
How to run SmolLM2-360M-Instruct
Run SmolLM2-360M-Instruct with vLLM
Generic example, not from the model's own docs: adjust flags (quantization, context length, parallelism) for your setup.
vllm serve HuggingFaceTB/SmolLM2-360M-Instruct --tensor-parallel-size 1Run SmolLM2-360M-Instruct with GGUF quantizations
Prebuilt GGUF weights published at unsloth/SmolLM2-360M-Instruct-GGUF. Run with llama.cpp's llama-server or load the repo directly in LM Studio.
llama-server -hf unsloth/SmolLM2-360M-Instruct-GGUFSource: https://huggingface.co/unsloth/SmolLM2-360M-Instruct-GGUF
Deploy SmolLM2-360M-Instruct on Aquanode
Aquanode has no one-click deploy template for SmolLM2-360M-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× RTX 3070 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 SmolLM2 models
- SmolLM2-360M (362M, BF16)
- SmolLM2-135M (135M, BF16)
- SmolLM2-135M-Instruct (135M, BF16)
- SmolLM2-1.7B (1.7B, BF16)
- SmolLM2-1.7B-Instruct (1.7B, BF16)
Lineage
Based on SmolLM2-360M.
Quantized and fine-tuned versions
Alternatives at this size
Other models for text-generation within about a third of SmolLM2-360M-Instruct's 362M parameters, from other model lines.
- Qwen2.5-0.5B-Instruct (494M, BF16)
- gemma-3-270m (268M, BF16)
- gpt2-medium (380M, F32)
- pythia-410m (506M, F16)
- LFM2.5-350M (354M, BF16)
More on SmolLM2-360M-Instruct
Fits on an 8 GB GPU at BF16: every model that fits in 8 GB.
Best chat and assistants models: how SmolLM2-360M-Instruct ranks against the rest.
Related reading: RTX 3070 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.
Related on Aquanode: GPU requirements for LLaMmlein_1B_prerelease, GPU requirements for SmolLM2-1.7B, GPU requirements for mamba-2.8b-hf, and GPU requirements for SmolLM-135M-Instruct.
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