About Inference Academy
Last updated November 2026
Running language models on your own hardware is one of the most useful skills in AI right now, and one of the worst-documented. Knowledge lives in scattered forum threads, vendor blogs with conflicting benchmark numbers, and tutorials that went stale when a flag was renamed.
Inference Academy is a structured, hands-on curriculum for local inference: nine tracks, interactive tools and labs, each lesson dated so you know how fresh it is.
Principles
- Estimates are labeled as estimates. We show the formula and tell you to measure on your own hardware.
- Free is genuinely useful. Core lessons, every calculator and the troubleshooting wiki are free.
- Privacy by default. Tools run in your browser; the playground runs models on your own GPU.
Contact
Configure NEXT_PUBLIC_CONTACT_EMAIL to show a contact address here.