Browser-Based Calculator Crunches the Real Power Cost of Self-Hosted LLMs vs Cloud APIs
paq85 · reddit · 2026-10-09
Reddit user paq85 released a local LLM electricity cost calculator answering the question self-hosters actually care about: the power bill.
- Plug in GPU power draw (W), prompt/decode tok/s, and your kWh price to get cost per hour, per request, per month, and per 1M input tokens, with and without KV cache.
- Ships with presets for RTX 4070 Ti Super, RTX 5090, and RTX 5090 eco; cached tokens are modeled at 1/100 the time, with a fixed realistic scenario of 60k input / 50k cached / 2k output at 90% uptime.
- Compares each profile against a reference API ($0.25/1M input, $1.20/1M output) to show whether self-hosting saves money.
- Runs fully in-browser, no sign-up, no uploads. Author's tip: compare per-request cost, not monthly bills.
More from Infra
- Open-source sfFFT delivers 3.8-6.1x speedup for FFT convolutions on DGX Spark GB10 — IgorCarron · 2026-10-09
- GPUs already within 2x of brain efficiency, and still beat human workers on energy — MikePFrank · 2026-10-09
- Do AI agents still need Kubernetes? Berlin event says yes, with agent-on-K8s cases — Al_Grigor · 2026-10-09
- Cloud Backlogs Hit $1.69T, CoreWeave Posts $2.58B Quarter as Inference Becomes the Battleground — FinanceYF5 · 2026-10-09
- NVIDIA Is AI's Central Bank: A100 Paper Citations Still Beat H100+H200 Combined — FinanceYF5 · 2026-10-09
- PartyKit shuts down free hosted platform 2.5 years after Cloudflare acquisition — threepointone · 2026-10-09