Cheapest box to run Gemma-3-27B at Q8_0
Google’s Gemma-3-27B is 27B 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.
Gemma-3-27B at Q8_0 needs 32GB.
Cheapest Config that fits: Mac mini, M5 Pro, 48GB. Fits, 12GB headroom. Est. 7.0 tok/s.
Best offer today: $2,499 at Apple — in stock, checked 10 hours ago.
Collecting price history — 2 of 90 days. No timing verdict yet.
Offers — Mac mini, M5 Pro, 48GB, 512GB
- Applelist pricein stockchecked 10 hours ago$2,499
- Mac mini, M6, 16GB$1,099Won't fit−19GB headroom
- Mac mini, M6, 24GB$1,299Won't fit−11GB headroom
- Mac mini, M6, 32GB$1,499Won't fit−3GB headroom
- Mac mini, M5 Pro, 24GB$1,699Won't fit−11GB headroom
- Mac mini, M5 Pro, 48GB$2,499Fits12GB headroomest. 7.0 tok/s
- Mac Studio, M5 Max, 36GB$2,499Tight1GB headroomest. 13.9 tok/s
- Mac mini, M5 Pro, 64GB$2,899Fits27GB headroomest. 7.0 tok/s
- GeForce RTX 5090, 32GB$2,999.99+hostWon't fit−1GB headroom
- Mac Studio, M5 Max, 48GB$3,099Fits12GB headroomest. 13.9 tok/s
- Mac Studio, M5 Max, 64GB$3,499Fits27GB headroomest. 13.9 tok/s
- Fits88GB headroomest. 6.2 tok/s
- Fits88GB headroomest. 6.2 tok/s
- Fits88GB headroomest. 13.9 tok/s
- Fits56GB headroomest. 27.2 tok/s
- Fits216GB headroomest. 27.2 tok/s
- Fits472GB headroomest. 27.2 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Won't fit | −19GB | — |
| Mac mini, M6 | 24GB | $1,299 | $54 | Won't fit | −11GB | — |
| Mac mini, M6 | 32GB | $1,499 | $47 | Won't fit | −3GB | — |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Won't fit | −11GB | — |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Fits | 12GB | 7.0 tok/s |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Tight | 1GB | 13.9 tok/s |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Fits | 27GB | 7.0 tok/s |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Won't fit | −1GB | — |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Fits | 12GB | 13.9 tok/s |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Fits | 27GB | 13.9 tok/s |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Fits | 88GB | 6.2 tok/s |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Fits | 88GB | 6.2 tok/s |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Fits | 88GB | 13.9 tok/s |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Fits | 56GB | 27.2 tok/s |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Fits | 216GB | 27.2 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 472GB | 27.2 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 = 27B × 1.063 B/param (8.5 bits) + 0.254 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.65 ÷ (27B active × 1.063 B/param) = 28.7 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
- Gemma-3-27BQ8_0, est. 7.0 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.