Cheapest box to run Qwen3-VL-8B-Instruct at Q8_0
Alibaba’s Qwen3-VL-8B-Instruct is 8B parameters, dense. 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-VL-8B-Instruct at Q8_0 needs 11GB.
Cheapest Config that fits: Mac mini, M6, 24GB. Fits, 10GB headroom. Est. 13.0 tok/s.
Best offer today: $1,299 at Expercom — in stock, checked 11 minutes ago.
Collecting price history — 3 of 90 days. No timing verdict yet.
Offers — Mac mini, M6, 24GB, 512GB
- Mac mini, M6, 16GB$1,099Tight2GB headroomest. 13.0 tok/s
- Mac mini, M6, 24GB$1,299Fits10GB headroomest. 13.0 tok/s
- Mac mini, M6, 32GB$1,499Fits18GB headroomest. 13.0 tok/s
- Mac mini, M5 Pro, 24GB$1,699Fits10GB headroomest. 23.5 tok/s
- Mac mini, M5 Pro, 48GB$2,499Fits33GB headroomest. 23.5 tok/s
- Mac Studio, M5 Max, 36GB$2,499Fits22GB headroomest. 47.0 tok/s
- Mac mini, M5 Pro, 64GB$2,899Fits48GB headroomest. 23.5 tok/s
- GeForce RTX 5090, 32GB$2,999.99+hostFits20GB headroomest. 137 tok/s
- Fits48GB headroomest. 41.8 tok/s
- Mac Studio, M5 Max, 48GB$3,099Fits33GB headroomest. 47.0 tok/s
- Mac Studio, M5 Max, 64GB$3,499Fits48GB headroomest. 47.0 tok/s
- Fits109GB headroomest. 20.9 tok/s
- Fits77GB headroomest. 62.6 tok/s
- Fits109GB headroomest. 41.8 tok/s
- Fits109GB headroomest. 20.9 tok/s
- Fits109GB headroomest. 47.0 tok/s
- Fits77GB headroomest. 91.8 tok/s
- Fits237GB headroomest. 62.6 tok/s
- Fits237GB headroomest. 91.8 tok/s
- Fits493GB headroomest. 91.8 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Tight | 2GB | 13.0 tok/s |
| Mac mini, M6 | 24GB | $1,299 | $54 | Fits | 10GB | 13.0 tok/s |
| Mac mini, M6 | 32GB | $1,499 | $47 | Fits | 18GB | 13.0 tok/s |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Fits | 10GB | 23.5 tok/s |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Fits | 33GB | 23.5 tok/s |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Fits | 22GB | 47.0 tok/s |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Fits | 48GB | 23.5 tok/s |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Fits | 20GB | 137 tok/s |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Fits | 48GB | 41.8 tok/s |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Fits | 33GB | 47.0 tok/s |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Fits | 48GB | 47.0 tok/s |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Fits | 109GB | 20.9 tok/s |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Fits | 77GB | 62.6 tok/s |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Fits | 109GB | 41.8 tok/s |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Fits | 109GB | 20.9 tok/s |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Fits | 109GB | 47.0 tok/s |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Fits | 77GB | 91.8 tok/s |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Fits | 237GB | 62.6 tok/s |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Fits | 237GB | 91.8 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 493GB | 91.8 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 = 8B × 1.063 B/param (8.5 bits) + 0.147 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.65 ÷ (8B active × 1.063 B/param) = 8.5 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 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.