Optimaiz field guides
Original guides about free LLM APIs, OpenAI-compatible local models, multi-machine routing, hybrid architecture, and the observability needed to trust the result.
Cost guide
7 min read · Updated Sep 15, 2026
A zero-dollar API bill is possible for local runs, but local inference still consumes hardware, electricity, and engineering time. Here is the honest comparison.
Read guide →Implementation guide
8 min read · Updated Sep 15, 2026
Keep the OpenAI SDK, change the base URL, select optimaiz.local, and let an approved computer run the request.
Read guide →Architecture
7 min read · Updated Sep 15, 2026
The useful question is not “local or cloud?” It is “which route is right for this workload under this policy?”
Read guide →Reliability
8 min read · Updated Sep 15, 2026
Treat laptops as an eligible compute pool, keep machine identity out of normal application calls, and make recovery rules explicit.
Read guide →Observability
7 min read · Updated Sep 15, 2026
Observability should explain not only what the model returned, but where it ran, why that route was chosen, and whether recovery changed cost or privacy.
Read guide →Start with the product
Install Optimaiz Local, approve an app, and make a standard OpenAI SDK call through your own machine.
