Every catalog hardware unit ranked by composite score (0–1000): measured tok/s, VRAM fit, ecosystem support, perf-per-watt. 2 of 154 ranks anchored to a measured benchmark — the rest are honestly flagged as extrapolated or estimated.
Methodology: /methodology · Run your own: curl -fsSL runlocalai.co/bench.mjs -o bench.mjs && node bench.mjs
22 units shown · sorted by score
| # | Hardware | Tier | Score | Data |
|---|---|---|---|---|
| 1 | NVIDIA GeForce RTX 5050 nvidia · entry · 8GB | D | 291 | Estimated |
| 2 | NVIDIA GeForce GTX 1080 nvidia · mid · 8GB | D | 286 | Estimated |
| 3 | NVIDIA GeForce RTX 4060 nvidia · entry · 8GB | D | 279 | Estimated |
| 4 | NVIDIA GeForce RTX 4060 Ti 8GB nvidia · mid · 8GB | D | 278 | Estimated |
| 5 | NVIDIA GeForce GTX 1070 nvidia · mid · 8GB | D | 270 | Estimated |
| 6 | NVIDIA GeForce GTX 1070 Ti nvidia · mid · 8GB | D | 268 | Estimated |
| 7 | NVIDIA GeForce RTX 3050 nvidia · entry · 8GB | D | 263 | Estimated |
| 8 | NVIDIA GeForce GTX 1660 Super nvidia · mid · 6GB | D | 261 | Estimated |
| 9 | ASUS ROG Strix Scar 18 (RTX 5090 Mobile) nvidia · enthusiast · 24GB | D | 259 | Estimated |
| 10 | Razer Blade 16 (2025, RTX 5090 Mobile) nvidia · enthusiast · 24GB | D | 259 | Estimated |
| 11 | NVIDIA GeForce RTX 2060 nvidia · mid · 6GB | D | 257 | Estimated |
| 12 | NVIDIA GeForce GTX 1660 Ti nvidia · mid · 6GB | D | 247 | Estimated |
| 13 | Lenovo Legion 5 Pro Gen 7 (RTX 3080 16GB) nvidia · high · 16GB | D | 238 | Estimated |
| 14 | NVIDIA GeForce RTX 3050 Ti (Mobile) nvidia · mobile · 4GB | D | 224 | Estimated |
| 15 | NVIDIA GeForce GTX 1650 Super nvidia · entry · 4GB | D | 221 | Estimated |
| 16 | NVIDIA GeForce GTX 1660 nvidia · mid · 6GB | D | 218 | Estimated |
| 17 | NVIDIA GeForce GTX 1060 3GB nvidia · entry · 3GB | D | 218 | Estimated |
| 18 | NVIDIA GeForce GTX 1060 6GB nvidia · entry · 6GB | D | 218 | Estimated |
| 19 | ASUS Ascent GX10 (NVIDIA GB10) nvidia · workstation | D | 217 | Estimated |
| 20 | NVIDIA GeForce GTX 1650 nvidia · entry · 4GB | D | 204 | Estimated |
| 21 | NVIDIA GeForce GTX 1050 Ti nvidia · entry · 4GB | D | 198 | Estimated |
| 22 | NVIDIA DGX Spark (Project Digits) nvidia · workstation | D | 140 | Estimated |
Amazon search links — we may earn a small commission at no extra cost to you. How we make money.
Steady-state tok/s on a representative 7B/8B Q4 model. Measured from real benchmark rows, or extrapolated from VRAM bandwidth × runtime-stack efficiency.
How comfortably the rig holds 7B / 32B / 70B class models. Apple unified memory counts; NPU/SoC system RAM counts.
CUDA / MLX / ROCm / Vulkan reach. Real-world friction the operator hits when installing tools.
Tok/s per watt. Mobile / NPU class scores well; dense desktop GPUs trade efficiency for absolute throughput.
A confidence multiplier (1.0 measured · 0.85 extrapolated · 0.7 estimated) discounts the headline so we don't pretend to know more than we do. Score is recomputed on every page load against the latest catalog + benchmark data — submit your own run with runlocalai-bench --submit --hardware your-rig to firm up the numbers.