Llama 3.3 70B on Mac mini M4 Pro 64 GB
70.554B on 64 GB at 273 GB/s. Hardware details · Model details
Yes. Llama 3.3 70B fits on Mac mini M4 Pro 64 GB at the recommended
MLX 4-bit configuration, requiring approximately 43.1 GB at 8K context. Practical context capacity is 32K. Expected decode for the recommended configuration is ~3.9–5.7 t/s. EstimatedCompatibilityComfortable
Calculated fit
Yes
Calculated Uses under 80% of usable memory. Room for a long context and other work.
Recommended quantization
MLX 4-bit
Calculated mlx
Memory required
43.1 GB
Calculated of 54.4 GB usable — 79%
Max practical context
32K
Calculated model supports 128K
Memory budget at 8K context
Model weights ⓘ
38.4 GB Calculated
KV cache ⓘ
2.5 GB Calculated
Runtime overhead ⓘ
2.2 GB Estimated
Total required
43.1 GB Calculated
Headroom ⓘ
11.3 GB Calculated
Highest-precision quantization that leaves headroom: uses 79% of usable memory at 8K context.
Unified memory: weights, KV cache and the OS share the same 64 GB pool. We assume 85% is addressable by the GPU, which matches a headless machine with the wired-memory limit raised.
Performance
Estimated performance · recommended MLX 4-bit
Decode, prefill and TTFT below are estimates for MLX 4-bit. We do not have a comparable MLX 4-bit measurement on this machine.
Estimated decode
~3.9–5.7 t/s
Estimated MLX 4-bit; calculated range
Estimated prefill
~17.2–35.8 t/s
Estimated MLX 4-bit; calculated range
Estimated TTFT at 8K
~228.9–476.4 s
Estimated MLX 4-bit; calculated range
Hardware load reference
62 W
Measured machine-level load; not this model run
How the estimate is calculated
- Decode: reading 70.6B active parameters at 4.50 bits/weight takes 207.67 ms at 273 GB/s x 70% achieved efficiency.
- Prefill: 17 TFLOPS (FP16) x 22% assumed model-FLOPs utilisation, divided by 2 x 70.6B parameters per token.
- Prefill throughput is highly engine-dependent. Flash attention, batch size and quantized KV all move this number substantially, and for sparse mixture-of-experts models it is the least reliable figure we produce.
- This is a calculated estimate, not a measurement. It assumes a single request, a short prompt, no speculative decoding and a warm model already resident in memory.
Estimator version estimator_v1. Stored with every estimated row so old estimates can be regenerated when the model improves.
Submit a benchmarkSubmissions stay pending until reviewed.
This model at other quantizations on this machine
| Quantization | Resident | Disk | Total needed | Utilisation | Fit | Max context |
|---|---|---|---|---|---|---|
MLX 4-bitrecommended | 38.4 GB | 38.4 GB | 43.1 GB | 79% | Comfortable | 32K |
Q4_K_M | 41.0 GB | 41.0 GB | 45.8 GB | 84% | Fits | 16K |
Q5_K_M | 48.1 GB | 48.1 GB | 53.1 GB | 98% | Does not fit | 4K |
Q8_0 | 71.2 GB | 71.2 GB | 76.8 GB | 141% | Does not fit | 0 |
BF16 | 131.4 GB | 131.4 GB | 138.9 GB | 255% | Does not fit | 0 |
Cost of running Llama 3.3 70B on this machineUnited States (federal) · C corporation · 8h/day
Monthly economic cost
$24
Calculated after tax
Codex/Claude Code
$100/mo
≈ $100/mo · local is $76 less
Net cash at purchase
$2,251
Calculated VAT not reclaimable
Total over 5 years
$1,414
Calculated after tax, after resale
Cost per USD/1M tokens
$38.81
Calculated 0.6M tokens/month
The $100 comparison uses the Codex Pro 5x / Claude Max 5x plans. A fixed planning conversion is used for USD. This compares monthly spend only: subscriptions have usage limits, local hardware has different capabilities and constraints, and taxes or regional pricing may change the charged amount. Prices checked 7 September 2026.
Purchase-price input: $2,251 in United States (federal), tax/VAT excluded · estimated · Apple Store NL (converted catalogue estimate). This is the localized purchase input; the after-tax economic cost is calculated separately below.
Monthly breakdown
| Depreciation 1,576 over 5 years, straight-line to a 675 residual | $26.26 |
| Electricity 2.7 kWh/month at 0.140/kWh | $0.38 |
| Cost of capital 4.0%/yr on 1,463 average capital employed | $4.88 |
| Monthly cost before tax | $31.53 |
| Electricity tax shield Running costs are deductible business expenses | −$0.08 |
| First-year expensing §179 (100.0%) | −$7.88 |
| Monthly economic cost after tax | $23.57 |
Three different numbers, deliberately
Cash cost
$2,251
Money that leaves the bank account on day one, net of reclaimable VAT.
Accounting depreciation
$26.26/month
$1,576 written down over 5 years to a $675 residual.
After-tax economic cost
$23.57/month
Depreciation plus running costs plus cost of capital, less the tax those deductions save. This is the figure to compare between machines.
Assumptions and sources (verified 2026-09-07)
VAT rate
0% Official spec
VAT recoverable
0% Official spec
Federal C-corporation estimate at the flat 21% rate. Pass-through entities should use the sole-proprietor estimate as a rougher proxy.
Effective deduction rate
21.00% Calculated
Headline marginal rate 21.00%.
Depreciation
5 years, straight-line Assumption
Residual value
$675 (30%) Assumption
Extrapolated from the three-year curve.
Electricity
$0.140/kWh Assumption
Average power draw
16 W Calculated
Load 62 W for 20% of powered hours, idle 4 W for the rest — a machine that is on is not generating tokens the whole time.
Investment allowances
§179 100.0% Official spec
Worth $473 in total — first-year expensing that replaces later tax depreciation.
Cost of capital
$4.88/month Assumption
- VAT rate: No US federal VAT · verified 2026-09-07
- Marginal tax rate: IRS Publication 542 — Corporations · verified 2026-09-07
- VAT recoverable fraction: site assumption · verified 2026-09-07
- Useful life: IRS Publication 946 — How To Depreciate Property · verified 2026-09-07
- Electricity price: Site assumption · verified 2026-09-07
Assumes the machine generates tokens 20% of its 176 powered hours per month. Cost per token scales inversely with this number — halve the utilisation and the cost per token doubles.
Federal planning estimate only, not tax advice. State and local income tax, sales/use tax and incentives are excluded. It assumes 100% business use, a Section 179 election, enough business income to use it, and that the full annual limit remains available.
Efficiency metrics
Decode per $1,000 spent
2.1 t/s
Calculated purchase price only
Decode per $100/month
20.3 t/s
Calculated after-tax ownership cost
Tokens per joule
0.31
Calculated same as tokens/s per watt
USD per 1M output tokens
$38.81
Calculated at 20% utilisation
Local versus hosted APIs
| Hosted model | USD/1M output | Break-even | API at your volume | Verdict | Comparison type |
|---|---|---|---|---|---|
| Qwen3.8 Flash Alibaba Cloud | $0.42 | 15.3M/mo | $1 | API cheaper | Different model |
| DeepSeek-V4.1 Flash DeepSeek | $0.60 | 13.1M/mo | $1 | API cheaper | Different model |
| DeepSeek-V4 Pro DeepSeek | $1.98 | 3.2M/mo | $4 | API cheaper | Different model |
| DeepSeek V4 Pro (Together) Together AI | $4.40 | 1.1M/mo | $13 | API cheaper | Different model |
| GLM-5.3 Z.ai | $4.40 | 1.5M/mo | $9 | API cheaper | Different model |
| Claude Haiku 4.5 Anthropic | $5.00 | 1.8M/mo | $8 | API cheaper | Different model |
Different-model comparison: the hosted model is not the model you would run locally. Treat this as a workload-quality trade-off, not a direct economic equivalence — the frontier model may complete a task in fewer tokens, or complete tasks the local model cannot. Break-even is the monthly output volume at which API spend equals the $24/month economic cost of owning this machine, assuming 8 input tokens per output token. Hosted prices are published in USD. This table is the canonical US default.
Llama 3.3 70B on other hardware
| Hardware | Decode | Price | Fit |
|---|---|---|---|
| NVIDIA DGX H200 (8x H200, 1,128 GB) | ~198–285 t/s | $359,430 * | Comfortable |
| Dual RTX 5090 workstation (2x 32 GB) | ~35.8–51.5 t/s | $5,854 * | Fits |
| Quad RTX 5090 workstation (4x 32 GB) | ~32.3–46.4 t/s | $12,611 * | Comfortable |
| Lenovo ThinkStation PX (4x RTX PRO 6000, 384 GB) | ~27.9–40.2 t/s | $58,553 * | Comfortable |
| RTX PRO 6000 Blackwell workstation (96 GB) | ~21.3–30.7 t/s | $10,359 * | Comfortable |
| RTX PRO 6000 Max-Q workstation (96 GB, 300 W) | ~21.3–30.7 t/s | $10,809 * | Comfortable |
Nearest alternatives to the Mac mini M4 Pro 64 GB