Tool · Run a model

Self-host vs API — at what volume does renting a GPU win?

Open-weight models can run on rented GPUs (~$1–$4/hr for an A100 / H100) or via per-token APIs. Plug in your numbers; we'll show the breakeven curve.

Monthly input tokens

30,000,000

@ 1,000,000/day

Monthly output tokens

15,000,000

@ 500,000/day

Self-host GPU

Nvidia B200

needs 179 GB VRAM

Self-host monthly

$2,008

24/7 on cheapest provider

API pricing — cheapest first

Monthly bill at 30,000,000 input + 15,000,000 output tokens.

Provider	$/M in	$/M out	Monthly	vs self-host
DeepSeek Platform	$0.7	$0.8	$33.0	−$1,975.8	Open →
OpenRouter	$0.7	$0.8	$33.0	−$1,975.8	Open →
DeepInfra	$0.7	$0.8	$33.0	−$1,975.8	Open →
Groq	$0.75	$0.99	$37.35	−$1,971.45	Open →
Together AI	$2.0	$2.0	$90.0	−$1,918.8	Open →

Cheapest fitting GPU for DeepSeek R1 Distill Llama 70B at FP16. Live at DeepInfra at $2.79/hr → $2,008/month at 24/7 utilisation.