Detect your hardware in one command
The quick scan reads your CPU, RAM and GPU specifications locally on Windows — then you paste the JSON here for a personalized compatibility analysis.
- No installation of llmfit, Python or Node.js
- No CanIHost desktop app or account
- No automatic upload — you copy and paste the JSON yourself
Step 1 — Run the command
Pick your operating system, copy the command, run it in PowerShell (Windows) or Terminal (macOS, Linux), and press Enter. It prints a JSON object with your hardware details.
$cpu=Get-CimInstance Win32_Processor|Select-Object -First 1;$cs=Get-CimInstance Win32_ComputerSystem;$os=Get-CimInstance Win32_OperatingSystem;$totalRam=[math]::Round($cs.TotalPhysicalMemory/1GB,2);$availableRam=[math]::Round(($os.FreePhysicalMemory*1KB)/1GB,2);$gpuName="";$gpuVram=0;$gpuAvailable=0;$gpuCount=0;$backend="CPU";$hasGpu=$false;$nvidia=Get-Command nvidia-smi -ErrorAction SilentlyContinue;if($nvidia){$d=@(nvidia-smi --query-gpu=name,memory.total,memory.free --format=csv,noheader,nounits);if($d.Count -gt 0){$p=$d[0]-split ",";$gpuName=$p[0].Trim();$gpuVram=[math]::Round(([double]$p[1].Trim())/1024,2);$gpuAvailable=[math]::Round(([double]$p[2].Trim())/1024,2);$gpuCount=$d.Count;$backend="CUDA";$hasGpu=$true}}else{$gpus=@(Get-CimInstance Win32_VideoController|Where-Object{$_.Name -notmatch "Microsoft Basic|Remote Display"});if($gpus.Count -gt 0){$g=$gpus|Sort-Object AdapterRAM -Descending|Select-Object -First 1;$gpuName=$g.Name;$gpuCount=$gpus.Count;$hasGpu=$true;if($g.AdapterRAM -and $g.AdapterRAM -gt 0){$gpuVram=[math]::Round([double]$g.AdapterRAM/1GB,2)};if($gpuName -match "AMD|Radeon"){$backend="AMD"}elseif($gpuName -match "Intel"){$backend="Intel"}else{$backend="GPU"}}};[ordered]@{total_ram_gb=$totalRam;available_ram_gb=$availableRam;cpu_cores=$cpu.NumberOfLogicalProcessors;cpu_name=$cpu.Name.Trim();has_gpu=$hasGpu;gpu_vram_gb=$gpuVram;gpu_available_gb=$gpuAvailable;gpu_name=$gpuName;gpu_count=$gpuCount;unified_memory=$false;backend=$backend}|ConvertTo-JsonThe command only reads hardware information (CPU name, thread count, RAM, GPU name and memory). It does not install software, modify anything, or send data anywhere.
On Apple Macs with unified memory, the total RAM is reported as the GPU memory — that is how local inference sees it.
Step 2 — Paste the JSON
Copy the JSON output from PowerShell, then use the scan tab in the checker below. The pasted JSON is used for the calculation and is not stored.
What the fields mean
- total_ram_gb / available_ram_gb
- System memory — total and currently free.
- cpu_cores
- Logical processor count (threads).
- gpu_vram_gb
- Dedicated GPU memory, via nvidia-smi when available.
Analyze your scan
Your system
Not sure about a value? A close estimate still produces a useful compatibility analysis.
Recommended for your PC
Prefer reading first? Browse the guides to understand what VRAM, quantization and fit levels mean before you check.
Prefer to enter specs manually?
The manual checker takes the same inputs by hand — useful on non-Windows machines or if you already know your specs.