Cheapest box to run Qwen3.5-35B-A3B at Q6_K
Alibaba’s Qwen3.5-35B-A3B is 35B parameters, 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.
Qwen3.5-35B-A3B at Q6_K needs 30GB.
Cheapest Config that fits: Mac mini, M5 Pro, 48GB. Fits, 14GB headroom. Est. 62.4 tok/s.
Best offer today: $2,499 at Apple — in stock, checked 32 minutes ago.
Price history: 34 days recorded.
A hybrid: only 10 of 40 layers are full attention. The Gated DeltaNet layers keep a constant-size state and cache nothing per token, so the K/V figure counts the full-attention layers alone.
Offers — Mac mini, M5 Pro, 48GB, 512GB
- Applelist pricein stockchecked 32 minutes ago$2,499
- Mac mini, M6, 16GB$1,099Won't fit−17GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−9GB headroom
- Mac mini, M6, 32GB$1,499Won't fit−1GB headroom
- Mac mini, M5 Pro, 24GB$1,699Won't fit−9GB headroom
- Mac mini, M5 Pro, 48GB$2,499Fits14GB headroomest. 62.4 tok/s
- Mac Studio, M5 Max, 36GB$2,499Tight3GB headroomest. 93.5 tok/s
- Mac mini, M5 Pro, 64GB$2,899Fits29GB headroomest. 62.4 tok/s
- GeForce RTX 5090, 32GB$2,999.99+hostTight1GB headroomest. 364 tok/s
- Fits29GB headroomest. 111 tok/s
- Mac Studio, M5 Max, 48GB$3,099Fits14GB headroomest. 125 tok/s
- Mac Studio, M5 Max, 64GB$3,499Fits29GB headroomest. 125 tok/s
- Fits90GB headroomest. 55.5 tok/s
- Fits58GB headroomest. 166 tok/s
- Fits90GB headroomest. 111 tok/s
- Fits90GB headroomest. 55.5 tok/s
- Fits90GB headroomest. 125 tok/s
- Fits58GB headroomest. 244 tok/s
- Fits218GB headroomest. 166 tok/s
- Fits218GB headroomest. 244 tok/s
- Fits474GB headroomest. 244 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −17GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −9GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Won't fit | −1GB | — |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −9GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Fits | 14GB | 62.4 tok/s |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Tight | 3GB | 93.5 tok/s |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Fits | 29GB | 62.4 tok/s |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Tight | 1GB | 364 tok/s |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Fits | 29GB | 111 tok/s |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Fits | 14GB | 125 tok/s |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Fits | 29GB | 125 tok/s |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Fits | 90GB | 55.5 tok/s |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Fits | 58GB | 166 tok/s |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Fits | 90GB | 111 tok/s |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Fits | 90GB | 55.5 tok/s |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Fits | 90GB | 125 tok/s |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Fits | 58GB | 244 tok/s |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Fits | 218GB | 166 tok/s |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Fits | 218GB | 244 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 474GB | 244 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 = 35B × 0.820 B/param (6.56 bits) + 0.021 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (3B active × 0.820 B/param) = 2.5 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
- Qwen3-Coder-NextQ3_K_M, est. 105 tok/s
- Llama-3.3-70BQ3_K_M, est. 5.8 tok/s
- Qwen3.5-35B-A3BQ8_0, est. 48.2 tok/s
- Qwen3-32BQ8_0, est. 5.9 tok/s
- Gemma-4-31BQ8_0, est. 6.1 tok/s
- Qwen3-30B-A3BQ8_0, est. 43.8 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.