Cheapest box to run Qwen3-30B-A3B at Q6_K
Alibaba’s Qwen3-30B-A3B is 30.5B parameters, 3.3B 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.
Qwen3-30B-A3B at Q6_K needs 27GB.
Cheapest Config that fits: Mac mini, M5 Pro, 48GB. Fits, 17GB headroom. Est. 56.7 tok/s.
Best offer today: $2,499 at Apple — in stock, checked 45 hours ago.
Collecting price history — 3 of 90 days. No timing verdict yet.
Offers — Mac mini, M5 Pro, 48GB, 512GB
- Applelist pricein stockchecked 45 hours ago$2,499
- Mac mini, M6, 16GB$1,099Won't fit−14GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−6GB headroom
- Mac mini, M6, 32GB$1,499Tight2GB headroomest. 31.4 tok/s
- Mac mini, M5 Pro, 24GB$1,699Won't fit−6GB headroom
- Mac mini, M5 Pro, 48GB$2,499Fits17GB headroomest. 56.7 tok/s
- Mac Studio, M5 Max, 36GB$2,499Fits6GB headroomest. 113 tok/s
- Mac mini, M5 Pro, 64GB$2,899Fits32GB headroomest. 56.7 tok/s
- GeForce RTX 5090, 32GB$2,999.99+hostTight4GB headroomest. 331 tok/s
- Fits32GB headroomest. 101 tok/s
- Mac Studio, M5 Max, 48GB$3,099Fits17GB headroomest. 113 tok/s
- Mac Studio, M5 Max, 64GB$3,499Fits32GB headroomest. 113 tok/s
- Fits93GB headroomest. 50.4 tok/s
- Fits61GB headroomest. 151 tok/s
- Fits93GB headroomest. 101 tok/s
- Fits93GB headroomest. 50.4 tok/s
- Fits93GB headroomest. 113 tok/s
- Fits61GB headroomest. 222 tok/s
- Fits221GB headroomest. 151 tok/s
- Fits221GB headroomest. 222 tok/s
- Fits477GB headroomest. 222 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −14GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −6GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Tight | 2GB | 31.4 tok/s |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −6GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Fits | 17GB | 56.7 tok/s |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Fits | 6GB | 113 tok/s |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Fits | 32GB | 56.7 tok/s |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Tight | 4GB | 331 tok/s |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Fits | 32GB | 101 tok/s |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Fits | 17GB | 113 tok/s |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Fits | 32GB | 113 tok/s |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Fits | 93GB | 50.4 tok/s |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Fits | 61GB | 151 tok/s |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Fits | 93GB | 101 tok/s |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Fits | 93GB | 50.4 tok/s |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Fits | 93GB | 113 tok/s |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Fits | 61GB | 222 tok/s |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Fits | 221GB | 151 tok/s |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Fits | 221GB | 222 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 477GB | 222 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 = 30.5B × 0.820 B/param (6.56 bits) + 0.098 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (3.3B active × 0.820 B/param) = 2.7 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 mini M5 Pro 48GB runs
- Llama-3.3-70BQ3_K_M, est. 5.8 tok/s
- Qwen3-32BQ8_0, est. 5.9 tok/s
- Qwen3-30B-A3BQ8_0, est. 43.8 tok/s
- Gemma-3-27BQ8_0, est. 7.0 tok/s
- Devstral-Small-2507Q8_0, est. 7.8 tok/s
- gpt-oss-20bBF16, est. 21.3 tok/s
Best quant that still fits at 8K context, largest model first. Full list on the Config page.
Bandwidth 307GB/s, usable memory 44GB of 48GB.