Guides
Task-first walkthroughs for the things people actually do with OpenLLM.
The rest of the docs explain what OpenLLM is. These guides are the opposite: each
one starts from something you want to do and walks you straight through it. They
assume you have finished the Quickstart: one provider
connected, one sk-llm-… key, one tool wired up.
Try and switch models
Pull any model from any provider into your coding tool, and swap it mid-thread without restarting.
Route by traffic and task
Ride out a provider's rate limits, and pin specific models to specific jobs like UI work.
Make agents cheaper and faster
Use semantic code search and persistent memory so agents read less and repeat less.
Generate images and video
Drive image and video models from the CLI, on the same key as your text models.
Track usage and savings
Watch every request in real time and see what you save against metered API pricing.
Use multiple devices
Run your account on a laptop, desktop, and server, up to three devices per subscription.