---
title: "Self-hosted LLM Inference Cost Calculator (Sep 2026)"
description: "Calculate GPU rental cost per million generated tokens from a sourced hourly price and your measured tokens per second; compare an API price."
canonical: https://vpsledger.com/tools/llm-inference-cost-calculator/
verified: 2026-10-23
format: text/markdown
html: https://vpsledger.com/tools/llm-inference-cost-calculator/
---

# Self-hosted LLM Inference Cost Calculator (Sep 2026)

> Markdown alternate of <https://vpsledger.com/tools/llm-inference-cost-calculator/>. Generated at build time from the
> same dataset as the HTML page; prices verified 2026-10-23. Data is a build-time
> snapshot — no live API, no auth. Machine entry point: <https://vpsledger.com/ai/>.

[GPU cloud](/gpu/) / Inference cost calculator

# LLM inference cost calculator

Choose a sourced GPU price or enter your own. Tokens/second must come from your own measured workload; no default throughput is assumed. This counts generated tokens only.

GPU price (USD/hour)

Measured tokens/second

Utilization (%)

API price (USD/1M tokens)

Rental hours/month

Enter a GPU price and measured throughput to calculate.

Formula: hourly price ÷ (tokens/second × 3,600 × utilization) × 1,000,000. Monthly rental assumes the entered hours are paid. Break-even volume = monthly rental ÷ API cost per million tokens. It only applies when the GPU can serve that volume at the measured speed; setup, input tokens, storage, CPU, electricity, downtime and engineering costs are excluded.

For video generation API estimates, see the [Seedance 2.5 cost calculator](/tools/seedance-2-5-api-cost-calculator/).

Affiliate links, if added later, will be marked sponsored.