Hardware for Local AI
58 configurations
| Name | Type | VRAM | Bandwidth | TDP | Price |
|---|---|---|---|---|---|
| NVIDIA RTX 5090 llm-inference · image-gen | gpu | 32 GB | 1792 GB/s | 575W | $1,999 |
| Mac Studio M4 Ultra large-model-inference · multi-model-loaded | mac | 192 GB | 819 GB/s | 270W | $4,699 |
| NVIDIA RTX 5080 llm-inference · image-gen | gpu | 16 GB | 960 GB/s | 360W | $999 |
| NVIDIA RTX 5070 Ti llm-inference · gaming | gpu | 16 GB | 896 GB/s | 300W | $749 |
| Intel Arc B580 llm-inference-entry · budget-build | gpu | 12 GB | 456 GB/s | 190W | $249 |
| NVIDIA RTX 5060 Ti llm-inference-entry · gaming-1080p | gpu | 16 GB | 448 GB/s | 180W | $429 |
| NVIDIA RTX 5070 gaming · llm-inference-entry | gpu | 12 GB | 672 GB/s | 250W | $549 |
| Apple Mac Mini M4 llm-inference-entry · development | mini-pc | 16 GB | 120 GB/s | 65W | $599 |
| NVIDIA RTX 4090 llm-inference · image-gen | gpu | 24 GB | 1008 GB/s | 450W | $1,599 |
| NVIDIA RTX 5060 gaming-1080p · llm-inference-small | gpu | 8 GB | 448 GB/s | 145W | $299 |
| Apple MacBook Pro 16" M4 Max llm-inference · development | laptop | 48 GB | 546 GB/s | 95W | $3,499 |
| NVIDIA RTX 3090 llm-inference · image-gen | gpu | 24 GB | 936 GB/s | 350W | $1,499 |
| AMD Radeon RX 9070 XT gaming · llm-inference | gpu | 16 GB | 640 GB/s | 260W | $549 |
| NVIDIA RTX 4070 Ti Super llm-inference · gaming | gpu | 16 GB | 672 GB/s | 285W | $799 |
| Apple Mac Mini M4 Pro llm-inference · development | mini-pc | 24 GB | 273 GB/s | 80W | $1,399 |
| Apple MacBook Pro 14" M4 Pro llm-inference · development | laptop | 24 GB | 273 GB/s | 70W | $1,999 |
| Framework Desktop development · llm-inference-entry | mini-pc | 16 GB | 120 GB/s | 65W | $899 |
| Apple MacBook Air M4 llm-inference-entry · development-mobile | laptop | 16 GB | 120 GB/s | 45W | $999 |
| AMD Radeon RX 9070 gaming · llm-inference-entry | gpu | 16 GB | 512 GB/s | 220W | $449 |
| NVIDIA RTX 4080 Super llm-inference · gaming | gpu | 16 GB | 736 GB/s | 320W | $999 |
| AMD Radeon RX 7900 XTX llm-inference · gaming | gpu | 24 GB | 960 GB/s | 355W | $999 |
| AMD Radeon RX 7800 XT gaming · llm-inference | gpu | 16 GB | 624 GB/s | 263W | $499 |
| Intel Arc B570 budget-build · gaming-1080p | gpu | 10 GB | 360 GB/s | 150W | $219 |
| AMD Radeon RX 7600 XT gaming-1080p · llm-inference-entry | gpu | 16 GB | 288 GB/s | 190W | $329 |
| NVIDIA RTX 4070 Super gaming · llm-inference-entry | gpu | 12 GB | 504 GB/s | 220W | $599 |
| MINISFORUM MS-A1 development · llm-inference-entry | mini-pc | 32 GB | 96 GB/s | 70W | $749 |
| AMD Radeon RX 7900 XT llm-inference · gaming | gpu | 20 GB | 800 GB/s | 315W | $899 |
| Beelink SER9 development · llm-inference-entry | mini-pc | 16 GB | 120 GB/s | 65W | $799 |
| Intel Arc A770 16GB llm-inference-entry · budget-build | gpu | 16 GB | 560 GB/s | 225W | $349 |
| NVIDIA RTX 4060 Ti 16GB llm-inference-entry · content-creation-budget | gpu | 16 GB | 288 GB/s | 165W | $499 |
| NVIDIA Jetson Orin NX 16GB edge-ai · robotics | edge | 16 GB | 102 GB/s | 25W | $679 |
| NVIDIA RTX 3090 Ti llm-inference · multi-gpu | gpu | 24 GB | 1008 GB/s | 450W | $1,999 |
| NVIDIA RTX 5050 gaming-1080p · entry-level | gpu | 8 GB | 320 GB/s | 130W | $249 |
| NVIDIA RTX 6000 Ada Generation llm-inference · professional | gpu | 48 GB | 960 GB/s | 300W | $6,800 |
| AMD Radeon RX 7700 XT gaming · llm-inference-entry | gpu | 12 GB | 432 GB/s | 245W | $449 |
| NVIDIA RTX 5000 Ada Generation llm-inference · professional | gpu | 32 GB | 768 GB/s | 250W | $3,600 |
| Apple MacBook Air M3 entry-level · development-mobile | laptop | 8 GB | 100 GB/s | 40W | $1,099 |
| GMKtec NucBox M7 development · entry-level | mini-pc | 16 GB | 80 GB/s | 54W | $599 |
| NVIDIA RTX 4070 gaming · llm-inference-entry | gpu | 12 GB | 504 GB/s | 200W | $549 |
| NVIDIA Jetson Orin Nano 8GB edge-ai · robotics | edge | 8 GB | 68 GB/s | 15W | $299 |
| NVIDIA RTX 4090 Laptop GPU gaming · llm-inference-entry | laptop | 16 GB | 576 GB/s | 150W | $1,999 |
| Raspberry Pi 5 edge-ai · development | edge | 0 GB | 0 GB/s | 15W | $80 |
| NVIDIA RTX 4060 gaming-1080p · entry-level | gpu | 8 GB | 272 GB/s | 115W | $299 |
| NVIDIA RTX 3060 12GB llm-inference-entry · entry-level | gpu | 12 GB | 360 GB/s | 170W | $329 |
| NVIDIA L40S llm-inference · professional | gpu | 48 GB | 864 GB/s | 350W | $10,200 |
| Intel Arc A750 gaming-1080p · entry-level | gpu | 8 GB | 512 GB/s | 225W | $289 |
| AMD Radeon RX 7600 gaming-1080p · entry-level | gpu | 8 GB | 288 GB/s | 165W | $269 |
| NVIDIA RTX 4060 Ti 8GB gaming-1080p · entry-level | gpu | 8 GB | 288 GB/s | 160W | $399 |
| NVIDIA RTX 3080 12GB gaming · llm-inference-entry | gpu | 12 GB | 912 GB/s | 350W | $799 |
| NVIDIA RTX A6000 llm-inference · professional | gpu | 48 GB | 768 GB/s | 300W | $4,650 |
| NVIDIA RTX 3080 Ti gaming · llm-inference-entry | gpu | 12 GB | 912 GB/s | 350W | $1,199 |
| NVIDIA RTX 3080 10GB gaming · entry-level | gpu | 10 GB | 760 GB/s | 320W | $699 |
| NVIDIA RTX 3070 Ti gaming · entry-level | gpu | 8 GB | 608 GB/s | 290W | $599 |
| NVIDIA RTX 3050 gaming-1080p · entry-level | gpu | 8 GB | 224 GB/s | 130W | $249 |
| NVIDIA Tesla T4 llm-inference-entry · budget-build | gpu | 16 GB | 320 GB/s | 70W | $2,499 |
| NVIDIA RTX 3070 gaming-1080p · entry-level | gpu | 8 GB | 448 GB/s | 220W | $499 |
| NVIDIA Tesla P40 llm-inference-entry · multi-gpu | gpu | 24 GB | 347 GB/s | 250W | $5,999 |
| NVIDIA RTX 2080 Ti entry-level · gaming | gpu | 11 GB | 616 GB/s | 250W | $1,199 |