Skip to content
LLM·TCO
Compare API cost Self-hosting Usage Data Compare costs
Home›Self-hosting

Data calculators

The dated, sourced reference dataset behind the calculators. Every price carries a verification date and a link to its source — and is overridable in every tool.

GPU TCO Calculator (Owned Hardware)Total cost of ownership for a self-hosted LLM GPU: capex amortization plus power and operational overhead, as monthly cost and cost per million tokens.GPU Cloud Cost CalculatorMonthly cost of a rented inference GPU from its hourly rate and utilization, with presets for popular GPU clouds.Self-Hosting Cost Per Token CalculatorConvert a GPU hourly rate and throughput (tokens/second) into a cost per million tokens, at any utilization.VRAM & Model-Fit CalculatorHow much GPU memory a model needs at fp16/int8/int4, and which GPUs can run it — parameters times bytes per parameter plus overhead.GPU Utilization Break-Even CalculatorThe utilization a rented GPU must reach to match a given monthly API spend — the practical test of whether self-hosting is worth it.

LLM·TCO

The real cost of LLM inference. Free, source-cited calculators for the cost of AI inference — API vs self-hosting.

Publisher: Redbit S.r.l.s.
Viale della Grande Muraglia 494, 00144 Roma, Italy
VAT IT15237911001

Tools

  • Compare
  • API cost
  • Self-hosting
  • Usage
  • Data

Learn

  • API vs self-hosting
  • Methodology
  • Guides
  • Sources

Site

  • About
  • Contact
  • Privacy policy
  • Cookie policy
  • Terms
  • Cookie preferences
© 2026 Redbit S.r.l.s. — All rights reserved. Cost estimates for informational purposes only; not financial advice.