Cheapest box to run Kimi-K2-Instruct at Q6_K
Moonshot’s Kimi-K2-Instruct is 1000B parameters, 32B of them read per token. The number below is computed from that and each vendor’s published specs, not asserted. Switch the quant or the context and it recomputes in place.
Kimi-K2-Instruct at Q6_K needs 822GB.
Nothing on the Grid runs Kimi-K2-Instruct at Q6_K today. The largest Config, Mac Studio, M5 Ultra, 512GB, is 318GB short. At it fits with 13GB headroom.
A trillion parameters. No Config we track holds it at any quant, and we list it so the arithmetic says so plainly.
- Mac mini, M6, 16GB$1,099Won't fit−809GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−801GB headroom
- Mac mini, M6, 32GB$1,499Won't fit−793GB headroom
- Mac mini, M5 Pro, 24GB$1,699Won't fit−801GB headroom
- Mac mini, M5 Pro, 48GB$2,499Won't fit−778GB headroom
- Mac Studio, M5 Max, 36GB$2,499Won't fit−789GB headroom
- Mac mini, M5 Pro, 64GB$2,899Won't fit−763GB headroom
- GeForce RTX 5090, 32GB$2,999.99+hostWon't fit−791GB headroom
- Won't fit−763GB headroom
- Mac Studio, M5 Max, 48GB$3,099Won't fit−778GB headroom
- Mac Studio, M5 Max, 64GB$3,499Won't fit−763GB headroom
- Won't fit−701GB headroom
- Won't fit−733GB headroom
- Won't fit−702GB headroom
- Won't fit−701GB headroom
- Won't fit−702GB headroom
- Won't fit−733GB headroom
- Won't fit−574GB headroom
- Won't fit−574GB headroom
- Won't fit−318GB headroom
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −809GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −801GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Won't fit | −793GB | — |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −801GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Won't fit | −778GB | — |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Won't fit | −789GB | — |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Won't fit | −763GB | — |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Won't fit | −791GB | — |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Won't fit | −763GB | — |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Won't fit | −778GB | — |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Won't fit | −763GB | — |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Won't fit | −701GB | — |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Won't fit | −733GB | — |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Won't fit | −702GB | — |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Won't fit | −701GB | — |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Won't fit | −702GB | — |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Won't fit | −733GB | — |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Won't fit | −574GB | — |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Won't fit | −574GB | — |
| Mac Studio, M5 Ultra | 512GB | — | — | Won't fit | −318GB | — |
GeForce RTX 5090 is a card, not a computer: every price shown for it is the cheapest listing we found plus $1,000 for the minimum host machine.
memory = 1000B × 0.820 B/param (6.56 bits) + 0.070 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (32B active × 0.820 B/param) = 26.2 GB read per token
usable = memory − operating system reserve (8% on a Mac, 6% on a DGX Spark, 4% on a card)
Bandwidth is the vendor’s published figure. Every speed is an estimate and says so. Methodology.