Meta · 109B total / 17B active · MoE
Llama 4 Scout 109B-A17B needs about 71 GB at Q4_K_M. 11 of the 17 priced rigs we track can hold it. Here's the cheapest, the best value and the fastest. Prices & stock checked 2026-10-23. Prices change daily and some retailers block our checker, so always confirm on the store page.
Ryzen AI Max+ 395, 128GB
Cheapest that fits. Lowest price we've seen on any tracked rig that holds Llama 4 Scout 109B-A17B at 4-bit (71 of 96 GB usable).
Ryzen AI Max+ 395, 128GB
Best value. Lowest price per token/sec among complete systems where it fits comfortably: about $191 per tok/s at 4-bit. Same machine as the cheapest pick.
96GB card
Fastest. Most memory bandwidth of the rigs that fit it (1,792 GB/s), so the quickest decode: ~106 tok/s at 4-bit.
GPU price is card only: you still need a host PC.
Check priceUse Q4_K_M on most rigs: it needs ~71 GB. Q8_0 would need ~123 GB, which doesn't fit comfortably at a usable speed on anything we track.
| Machine | Price | 4-bit fit & speed | Suggested quant | $ per tok/s |
|---|---|---|---|---|
| Framework Desktop DIYStrix Halo | $3,449 | Fits~18 tok/s | Q4_K_M | $191 |
| GMKtec EVO-X2Strix Halo | $3,500 | Fits~18 tok/s | Q4_K_M | $194 |
| Minisforum MS-S1 MaxStrix Halo | $3,919 | Fits~18 tok/s | Q4_K_M | $217 |
| Beelink GTR9 ProStrix Halo | $4,349 | Fits~18 tok/s | Q4_K_M | $241 |
| Mac Studio M5 UltraMac Studio | $5,499 | Tight fit~60 tok/s | Q4_K_M | $92 |
| Lenovo ThinkStation PGXDGX Spark / GB10 | $5,859 | Fits~19 tok/s | Q4_K_M | $307 |
| ASUS Ascent GX10DGX Spark / GB10 | $5,999 | Fits~19 tok/s | Q4_K_M | $315 |
| NVIDIA DGX SparkDGX Spark / GB10 | $6,950 | Fits~19 tok/s | Q4_K_M | $364 |
| HP ZGX Nano G1nDGX Spark / GB10 | $7,699 | Fits~19 tok/s | Q4_K_M | $404 |
| Gigabyte AI TOP ATOMDGX Spark / GB10 | $8,499 | Fits~19 tok/s | Q4_K_M | $446 |
| PNY NVIDIA RTX PRO 6000 BlackwellRTX PRO 6000 (card only, needs a host PC) | $17,899 | Fits~106 tok/s | Q4_K_M | $169 |
Sorted by lowest price. Speeds are estimated single-user decode at ~8K context. Full fit table, including rigs that don't fit: Llama 4 Scout 109B-A17B memory requirements. How we estimate
About 71 GB at Q4_K_M and 123 GB at Q8_0, including runtime overhead and ~8K tokens of context for one user. Longer context needs more.
Of the rigs we track, the Framework Desktop DIY at $3,449, estimated ~18 tok/s at 4-bit. Prices checked 2026-10-23.
Q4_K_M is the default: about 71 GB. Q8_0 needs about 123 GB; use it only if you have the room and still get usable speed.
It has 17B active parameters per token, so speed is set by memory bandwidth: from ~18 tok/s on the slowest rig that fits to ~106 tok/s on the PNY NVIDIA RTX PRO 6000 Blackwell. Estimates, not benchmarks.
One short email a week: what got cheaper, what's back in stock, and which new open models your machine can run. Unsubscribe anytime.
