Cheapest box to run Gemma-3n-E4B at Q3_K_M
Google’s Gemma-3n-E4B is 6.8B 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.
Gemma-3n-E4B at Q3_K_M needs 5GB.
Cheapest Config that fits: Mac mini, M6, 16GB. Fits, 8GB headroom. Est. 56.5 tok/s.
Best offer today: $1,099 at Apple — in stock, checked 35 hours ago.
Collecting price history — 3 of 90 days. No timing verdict yet.
Google names this model for its 4B effective parameters, but 6.8B raw weights are what you load, so the memory arithmetic uses 6.8B. The K/V figure accounts for layer sharing and sliding-window attention and is an estimate.
Offers — Mac mini, M6, 16GB, 512GB
- Applelist pricein stockchecked 35 hours ago$1,099
- Mac mini, M6, 16GB$1,099Fits8GB headroomest. 56.5 tok/s
- Mac mini, M6, 24GB$1,299Fits16GB headroomest. 56.5 tok/s
- Mac mini, M6, 32GB$1,499Fits24GB headroomest. 56.5 tok/s
- Mac mini, M5 Pro, 24GB$1,699Fits16GB headroomest. 102 tok/s
- Mac mini, M5 Pro, 48GB$2,499Fits39GB headroomest. 102 tok/s
- Mac Studio, M5 Max, 36GB$2,499Fits28GB headroomest. 204 tok/s
- Mac mini, M5 Pro, 64GB$2,899Fits54GB headroomest. 102 tok/s
- GeForce RTX 5090, 32GB$2,999.99+hostFits26GB headroomest. 596 tok/s
- Fits54GB headroomest. 182 tok/s
- Mac Studio, M5 Max, 48GB$3,099Fits39GB headroomest. 204 tok/s
- Mac Studio, M5 Max, 64GB$3,499Fits54GB headroomest. 204 tok/s
- Fits115GB headroomest. 90.8 tok/s
- Fits83GB headroomest. 272 tok/s
- Fits115GB headroomest. 182 tok/s
- Fits115GB headroomest. 90.8 tok/s
- Fits115GB headroomest. 204 tok/s
- Fits83GB headroomest. 399 tok/s
- Fits243GB headroomest. 272 tok/s
- Fits243GB headroomest. 399 tok/s
- Fits499GB headroomest. 399 tok/s
| Config | Memory | Price | $/GB | Fit | Headroom | Est. speed |
|---|---|---|---|---|---|---|
| Mac mini, M6 | 16GB | $1,099 | $69 | Fits | 8GB | 56.5 tok/s |
| Mac mini, M6 | 24GB | $1,299 | $54 | Fits | 16GB | 56.5 tok/s |
| Mac mini, M6 | 32GB | $1,499 | $47 | Fits | 24GB | 56.5 tok/s |
| Mac mini, M5 Pro | 24GB | $1,699 | $71 | Fits | 16GB | 102 tok/s |
| Mac mini, M5 Pro | 48GB | $2,499 | $52 | Fits | 39GB | 102 tok/s |
| Mac Studio, M5 Max | 36GB | $2,499 | $69 | Fits | 28GB | 204 tok/s |
| Mac mini, M5 Pro | 64GB | $2,899 | $45 | Fits | 54GB | 102 tok/s |
| GeForce RTX 5090 | 32GB | $2,999.99+host | $94 | Fits | 26GB | 596 tok/s |
| Mac Studio, M4 Max, refurb | 64GB | $3,049 | $48 | Fits | 54GB | 182 tok/s |
| Mac Studio, M5 Max | 48GB | $3,099 | $65 | Fits | 39GB | 204 tok/s |
| Mac Studio, M5 Max | 64GB | $3,499 | $55 | Fits | 54GB | 204 tok/s |
| Ascent GX10, GB10 Grace Blackwell | 128GB | $3,999.99 | $31 | Fits | 115GB | 90.8 tok/s |
| Mac Studio, M3 Ultra, refurb | 96GB | $4,329 | $45 | Fits | 83GB | 272 tok/s |
| Mac Studio, M4 Max, refurb | 128GB | $4,409 | $34 | Fits | 115GB | 182 tok/s |
| DGX Spark, GB10 Grace Blackwell | 128GB | $4,699 | $37 | Fits | 115GB | 90.8 tok/s |
| Mac Studio, M5 Max | 128GB | $5,099 | $40 | Fits | 115GB | 204 tok/s |
| Mac Studio, M5 Ultra | 96GB | $5,499 | $57 | Fits | 83GB | 399 tok/s |
| Mac Studio, M3 Ultra, refurb | 256GB | $7,729 | $30 | Fits | 243GB | 272 tok/s |
| Mac Studio, M5 Ultra | 256GB | $9,499 | $37 | Fits | 243GB | 399 tok/s |
| Mac Studio, M5 Ultra | 512GB | — | — | Fits | 499GB | 399 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 = 6.8B × 0.489 B/param (3.91 bits) + 0.072 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.65 ÷ (4B active × 0.489 B/param) = 2.0 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 16GB runs
- gpt-oss-20bQ3_K_M, est. 48.3 tok/s
- Phi-4Q5_K_M, est. 11.1 tok/s
- Qwen3-VL-8B-InstructQ8_0, est. 13.0 tok/s
- Gemma-3n-E4BQ8_0, est. 26.0 tok/s
- Qwen3-4BBF16, est. 13.8 tok/s
Best quant that still fits at 8K context, largest model first. Full list on the Config page.
Bandwidth 170GB/s, usable memory 13GB of 16GB.