Home/Calculator
LLM memory
calculator.
Change the deployment assumptions and see exactly which part of the workload consumes VRAM.
Deployment worksheetMODEL / GPU FIT
Calculated estimateRecommended VRAM
8.6 GiBper GPU × 1
Includes 15% operating headroom for the selected runtime profile.
Deployment next stepCheck live capacity on Flatkey Compute.
Continue to Flatkey Compute ↗Carry this workload estimate into the product and review currently available GPU options.
Estimated allocation per GPU7.3 GiB each
Weights 4.3 GiBKV cache 1.0 GiBRuntime 2.0 GiB
Example cloud fitsBrowse profiles →
011× NVIDIA T416 GB each · Turing
Comfortable
$0.20sample / hour
021× RTX 409024 GB each · Ada Lovelace
Comfortable
$0.32sample / hour
031× NVIDIA L424 GB each · Ada Lovelace
Comfortable
$0.43sample / hour
Sample planning rates only—not live Flatkey quotes or availability.
Command template · vLLM
vllm serve YOUR_COMPATIBLE_QUANTIZED_MODEL_ID --tensor-parallel-size 1 --max-model-len 8192 --quantization YOUR_MATCHING_BACKEND4-bit is a memory assumption, not a checkpoint choice. Replace both placeholders with a compatible quantized artifact and its vLLM backend.
Planning estimate
Runtime kernels, allocator behavior, framework versions, multimodal inputs, speculative decoding, and provider-reserved memory can move the real number. Validate on the exact stack before committing capacity.
Read the methodology →