VRAM Calculator
Pick your GPU and what you use models for. Every open-weight model the site knows is sized for your card, graded on fit, given a speed estimate from the card's memory bandwidth, and ranked by how well it matches the job. Where community benchmarks exist on the same card, the measured number replaces the estimate.
8 GB VRAM·32 GB system RAM·256 GB/s·5.8 TFLOPS fp16
Balanced: quality first, then speed.
1 perfect·26 good·1 marginal·15 too tight
Sort
Memory assumes an F16 KV cache at 32k context; Perfect is under 60% of the pool, Good under 85%, Marginal under 98%. Offload paths cap at Good. Speed is the card's memory bandwidth divided by the bytes read per token, at 55% efficiency, as llmfit estimates it; a community median at the same quant replaces it. Click a row for the breakdown.