Cheapest box to run GLM-4.6 at Q8_0
Z.ai’s GLM-4.6 is 357B 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.
GLM-4.6 at Q8_0 needs 384GB.
Cheapest Config that fits: Mac Studio, M5 Ultra, 512GB. Fits, 120GB headroom. Est. 17.6 tok/s.
No price yet: Ships late October 2026. Apple has not published a price.
Collecting price history — 3 of 90 days. No timing verdict yet.
Offers — Mac Studio, M5 Ultra, 512GB, 1TB
Ships late October 2026. Apple has not published a price.
- Mac mini, M6, 16GB$1,099Won't fit−371GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−363GB headroom
- Mac mini, M6, 32GB$1,499Won't fit−355GB headroom
- Mac mini, M5 Pro, 24GB$1,699Won't fit−363GB headroom
- Mac mini, M5 Pro, 48GB$2,499Won't fit−339GB headroom
- Mac Studio, M5 Max, 36GB$2,499Won't fit−351GB headroom
- Mac mini, M5 Pro, 64GB$2,899Won't fit−325GB headroom
- GeForce RTX 5090, 32GB$2,999.99+hostWon't fit−353GB headroom
- Won't fit−325GB headroom
- Mac Studio, M5 Max, 48GB$3,099Won't fit−339GB headroom
- Mac Studio, M5 Max, 64GB$3,499Won't fit−325GB headroom
- Won't fit−263GB headroom
- Won't fit−295GB headroom
- Won't fit−264GB headroom
- Won't fit−263GB headroom
- Won't fit−264GB headroom
- Won't fit−295GB headroom
- Won't fit−136GB headroom
- Won't fit−136GB headroom
- Fits120GB headroomest. 17.6 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −371GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −363GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Won't fit | −355GB | — |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −363GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Won't fit | −339GB | — |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Won't fit | −351GB | — |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Won't fit | −325GB | — |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Won't fit | −353GB | — |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Won't fit | −325GB | — |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Won't fit | −339GB | — |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Won't fit | −325GB | — |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Won't fit | −263GB | — |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Won't fit | −295GB | — |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Won't fit | −264GB | — |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Won't fit | −263GB | — |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Won't fit | −264GB | — |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Won't fit | −295GB | — |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Won't fit | −136GB | — |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Won't fit | −136GB | — |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 120GB | 17.6 tok/s |
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 = 357B × 1.063 B/param (8.5 bits) + 0.377 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (32B active × 1.063 B/param) = 34.0 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.
What else Mac Studio M5 Ultra 512GB runs
- Kimi-K2-InstructQ3_K_M, est. 38.4 tok/s
- DeepSeek-V3Q5_K_M, est. 22.9 tok/s
- DeepSeek-R1Q5_K_M, est. 22.9 tok/s
- Qwen3-Coder-480B-A35B-InstructQ6_K, est. 20.9 tok/s
- Llama-4-Maverick-17B-128EQ8_0, est. 33.2 tok/s
- GLM-4.6Q8_0, est. 17.6 tok/s
Best quant that still fits at 8K context, largest model first. Full list on the Config page.
Bandwidth 1.2TB/s, usable memory 504GB of 512GB.