Used RTX 4090 workstation (24 GB)

Discontinued but abundant on the used market, and still the best value per GB/s if you can live with 24 GB. The reference point for 'is a used build cheaper than a new appliance'.

NVIDIAcomponentAvailable14 models fit1 measured
Memory
24 GB
Official spec GDDR6X
Memory bandwidth
1,008 GB/s
Official spec ~80% achieved in practice
Specifications
CPU
AMD Ryzen 7 7700X Official spec
GPU
GeForce RTX 4090 24 GB (used) Official spec
Memory
24 GB GDDR6X Official spec
Memory bandwidth
1,008 GB/s Official spec source
Achieved bandwidth ⓘ
~80% (806 GB/s) Measured
FP16 compute ⓘ
~165 TFLOPS Official spec
System RAM
64 GB Official spec
Architecture
Discrete GPU with dedicated VRAM Official spec
Nominal power
600 W Official spec
Measured load power
520 W Measured
Idle power
60 W Measured
Released
12 Oct 2022 Official spec
Representative used build. Price is a Marktplaats/eBay-class observation, not a retail price, and used-market prices move fast. Depreciation on an already-depreciated asset is much flatter, which changes the economics substantially.
Model performance on Used RTX 4090 workstation (24 GB)1 measured, 13 estimated
14 of 14 rows
Model↕QuantMemory↕Decode▼Prefill↕ContextFitConfidence↕
Nemotron 3.5 Lightning 30B-A3B
NVIDIA · 31.578B (3B active)
Q4_K_M
recommended
20.0 GB~237–342 t/s~3530–7320 t/s192KBorderlineEstimated
gpt-oss-20b
OpenAI · 20.915B (3.6B active)
MXFP4
recommended
13.0 GB~228–329 t/s~7910–16430 t/s128KComfortableEstimated
Qwen3-Coder 30B-A3B
Alibaba Qwen · 30.532B (3.3B active)
Q4_K_M
recommended
20.0 GB~221–317 t/s~3420–7100 t/s16KBorderlineEstimated
Qwen3 30B-A3B
Alibaba Qwen · 30.532B (3.3B active)
Q4_K_M
recommended
20.0 GB~221–317 t/s~3420–7100 t/s16KBorderlineEstimated
Gemma 4 26B-A4B
Google DeepMind · 25.806B (3.8B active)
Q4_K_M
recommended
18.3 GB~197–284 t/s~3470–7200 t/s16KFitsEstimated
Ternary Bonsai 2 27B
Prism ML · 27.36B
PTQ1_0
recommended
measured: PQ2_0
7.2 GB91.1 t/s1650 t/s192KComfortableLow
Qwen3 8B
Alibaba Qwen · 8.191B
Q8_0
recommended
10.6 GB~70.7–102 t/s~4190–8700 t/s40KComfortableEstimated
Phi-4 14B
Microsoft · 14.66B
Q5_K_M
recommended
12.9 GB~59.7–85.9 t/s~2340–4860 t/s16KComfortableEstimated
Devstral Small 24B
Mistral AI · 23.572B
Q4_K_M
recommended
16.4 GB~44.3–63.7 t/s~1460–3020 t/s32KComfortableEstimated
Mistral Small 3.2 24B
Mistral AI · 24.011B
Q4_K_M
recommended
16.6 GB~43.5–62.6 t/s~1430–2970 t/s32KComfortableEstimated
Gemma 3 27B
Google DeepMind · 27.432B
Q4_K_M
recommended
21.3 GB~38.3–55.1 t/s~1250–2600 t/s8KBorderlineEstimated
Qwen3.8 27B
Alibaba Qwen · 27.781B
Q4_K_M
recommended
18.1 GB~37.8–54.4 t/s~1240–2570 t/s32KFitsEstimated
Qwen3.6 27B
Alibaba Qwen · 27.781B
Q4_K_M
recommended
18.1 GB~37.8–54.4 t/s~1240–2570 t/s32KFitsEstimated
Granite 4.2 30B
IBM · 29.277B
Q4_K_M
recommended
20.5 GB~35.9–51.7 t/s~1170–2430 t/s8KBorderlineEstimated
Rows in grey are estimates from our bandwidth model, not measurements — they are shown as a range and never as a precise figure. Use the “measured only” filter to see just the 1 pairing on this machine that a real benchmark backs.
Submit a benchmarkContributions are reviewed before publication.
Ownership economicsUnited States (federal) · C corporation · 8h/day
Monthly economic cost
$27
Calculated after tax
Codex/Claude Code
$100/mo
≈ $100/mo · local is $73 less
Net cash at purchase
$2,071
Calculated VAT not reclaimable
Total over 5 years
$1,611
Calculated after tax, after resale
Cost per USD/1M tokens
$0.73
Calculated 36.7M 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: $2,071 in United States (federal), tax/VAT excluded · estimated · Used market (Marktplaats/eBay) (converted catalogue estimate). This is the localized purchase input; the after-tax economic cost is calculated separately below.

Monthly breakdown

Depreciation
1,615 over 5 years, straight-line to a 456 residual
$26.92
Electricity
26.8 kWh/month at 0.140/kWh
$3.75
Cost of capital
4.0%/yr on 1,263 average capital employed
$4.21
Monthly cost before tax$34.88
Electricity tax shield
Running costs are deductible business expenses
−$0.79
First-year expensing
§179 (100.0%)
−$7.25
Monthly economic cost after tax$26.84

Three different numbers, deliberately

Cash cost
$2,071
Money that leaves the bank account on day one, net of reclaimable VAT.
Accounting depreciation
$26.92/month
$1,615 written down over 5 years to a $456 residual.
After-tax economic cost
$26.84/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
$456 (22%) Assumption
Extrapolated.
Electricity
$0.140/kWh Assumption
Average power draw
152 W Calculated
Load 520 W for 20% of powered hours, idle 60 W for the rest — a machine that is on is not generating tokens the whole time.
Investment allowances
§179 100.0% Official spec
Worth $435 in total — first-year expensing that replaces later tax depreciation.
Cost of capital
$4.21/month Assumption
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.