GPU Value Finder
Rank GPUs by dollars per GB of VRAM for your budget and model size — the cheapest path to the VRAM you need.
- Best for
- best value GPU for local LLM
- Main output
- GPUs ranked by memory per dollar
Local AI Tools
Pick one focused calculator or planner for the decision in front of you: memory fit, hardware comparison, operating cost, platform choice, used-GPU risk, power planning, parts sizing, or workload matching.
Comparison
Comparison tools
Mini-AI-PC Finder
Pick the model you want to run and see which unified-memory mini PC (Strix Halo, DGX Spark, Framework, Mac) can run it — ranked cheapest-first.
- Best for
- which mini pc can run a 70b model
- Main output
- Mini PCs that run your model, cheapest first
Hardware Comparison
Compare two GPUs, Macs, or multi-GPU tiers on memory, planning price, power, and comfort zone.
- Best for
- local AI hardware comparison tool
- Main output
- Price-band comparison
Memory
Memory tools
VRAM Calculator
Estimate model memory needs from parameters, quantization, and context length before choosing hardware.
- Best for
- VRAM calculator for local LLMs
- Main output
- Estimated VRAM
VRAM + RAM Offload Calculator
Estimate what fits across VRAM and system RAM, an approximate llama.cpp layer split, and the likely offload severity.
- Best for
- how many layers to offload to GPU llama.cpp
- Main output
- Estimated full-GPU, offload, or no-fit status
Fine-Tuning VRAM Calculator
Estimate full, LoRA, or QLoRA memory with explicit assumptions and a conservative GPU reserve.
- Best for
- fine-tuning VRAM calculator
- Main output
- Conservative estimated-fit status
Cost
Cost tools
Electricity Cost Calculator
Estimate monthly and yearly power cost from GPU preset, runtime, utility rate, or custom watt draw.
- Best for
- local AI electricity cost calculator
- Main output
- Planning watts
Local vs Cloud Break-Even
Compare hardware cost plus electricity against a recurring cloud or subscription AI bill.
- Best for
- local AI vs cloud break-even calculator
- Main output
- Local power cost per month
Platform
Platform tools
Apple Silicon vs Discrete GPU
Choose the cleaner platform path from CUDA needs, model size, noise sensitivity, and workflow priority.
- Best for
- Apple Silicon vs discrete GPU for local AI
- Main output
- Platform recommendation
Buying Risk
Buying Risk tools
Used GPU Buying Checklist
Score listing, return-window, stress-test, VRAM-test, power, fit, and value risks before buying used.
- Best for
- used GPU buying checklist for local AI
- Main output
- Purchase risk level
Build Planning
Build Planning tools
Power, Thermals, and Noise Planner
Turn hardware wattage into PSU target, heat output, thermal risk, and noise guidance.
- Best for
- local AI power thermals noise planner
- Main output
- System planning watts
AI Rig Parts Planner
Size CPU, RAM, PSU, and storage around the GPU or unified-memory tier you plan to use.
- Best for
- AI rig parts planner
- Main output
- CPU target
AI Workload Chooser
Start with the job: coding, quiet Mac use, image generation, large LLMs, or budget experimentation.
- Best for
- AI workload chooser for local hardware
- Main output
- Recommended starting point
How To Choose
Use the narrowest tool that matches the decision
The tool pages are separated by reader job, not by keyword variants. Use the VRAM calculator for model memory, the electricity calculator for operating cost, the break-even calculator for ownership payback, and the build planners when the physical rig or workload path is the real question.