Cheapest box to run Qwen3-30B-A3B at Q3_K_M
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 Q3_K_M needs 17GB.
Cheapest Config that fits: Mac mini, M6, 24GB. Fits, 4GB headroom. Est. 52.7 tok/s.
Best offer today: $1,299 at Best Buy — in stock, checked 45 hours ago.
Collecting price history — 3 of 90 days. No timing verdict yet.
Offers — Mac mini, M6, 24GB, 512GB
- Mac mini, M6, 16GB$1,099Won't fit−4GB headroom
- Mac mini, M6, 24GB$1,299Fits4GB headroomest. 52.7 tok/s
- Mac mini, M6, 32GB$1,499Fits12GB headroomest. 52.7 tok/s
- Mac mini, M5 Pro, 24GB$1,699Fits4GB headroomest. 95.2 tok/s
- Mac mini, M5 Pro, 48GB$2,499Fits27GB headroomest. 95.2 tok/s
- Mac Studio, M5 Max, 36GB$2,499Fits16GB headroomest. 190 tok/s
- Mac mini, M5 Pro, 64GB$2,899Fits42GB headroomest. 95.2 tok/s
- GeForce RTX 5090, 32GB$2,999.99+hostFits14GB headroomest. 556 tok/s
- Fits42GB headroomest. 169 tok/s
- Mac Studio, M5 Max, 48GB$3,099Fits27GB headroomest. 190 tok/s
- Mac Studio, M5 Max, 64GB$3,499Fits42GB headroomest. 190 tok/s
- Fits103GB headroomest. 84.6 tok/s
- Fits71GB headroomest. 254 tok/s
- Fits103GB headroomest. 169 tok/s
- Fits103GB headroomest. 84.6 tok/s
- Fits103GB headroomest. 190 tok/s
- Fits71GB headroomest. 372 tok/s
- Fits231GB headroomest. 254 tok/s
- Fits231GB headroomest. 372 tok/s
- Fits487GB headroomest. 372 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −4GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Fits | 4GB | 52.7 tok/s |
| Mac mini, M6 | 32GB | $1,499 | $47 | Fits | 12GB | 52.7 tok/s |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Fits | 4GB | 95.2 tok/s |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Fits | 27GB | 95.2 tok/s |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Fits | 16GB | 190 tok/s |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Fits | 42GB | 95.2 tok/s |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Fits | 14GB | 556 tok/s |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Fits | 42GB | 169 tok/s |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Fits | 27GB | 190 tok/s |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Fits | 42GB | 190 tok/s |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Fits | 103GB | 84.6 tok/s |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Fits | 71GB | 254 tok/s |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Fits | 103GB | 169 tok/s |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Fits | 103GB | 84.6 tok/s |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Fits | 103GB | 190 tok/s |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Fits | 71GB | 372 tok/s |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Fits | 231GB | 254 tok/s |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Fits | 231GB | 372 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 487GB | 372 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 × 0.489 B/param (3.91 bits) + 0.098 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (3.3B active × 0.489 B/param) = 1.6 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 M6 24GB runs
- Qwen3-32BQ3_K_M, est. 7.1 tok/s
- Qwen3-30B-A3BQ4_K_M, est. 42.7 tok/s
- Gemma-3-27BQ4_K_M, est. 6.8 tok/s
- Devstral-Small-2507Q5_K_M, est. 6.5 tok/s
- gpt-oss-20bQ6_K, est. 28.8 tok/s
- Phi-4Q8_0, est. 7.4 tok/s
Best quant that still fits at 8K context, largest model first. Full list on the Config page.
Bandwidth 170GB/s, usable memory 21GB of 24GB.