LLM Resource Calculator
Calculate hardware requirements, power consumption, and performance metrics for LLM models
Number of Parameters (billions)
Quantization Method
No Quantization
INT8
INT4
QLoRA 4-bit
GPTQ 4-bit
Model Precision
Float16
Float32
Int8
Batch Size
Context Length
GPU Model
NVIDIA A100 (80GB)
NVIDIA A6000
NVIDIA RTX 4090
NVIDIA RTX 3090
Memory Requirements
Model Size:
-
Total RAM Required:
-
Disk Space:
-
Performance Metrics
Tokens per Second:
-
Power Consumption:
-
Energy per Token:
-