Practical overview
When it fits
Use yzma when building a Go application that should perform language-model inference locally through llama.cpp, rather than relying on a separate model server.
Who it is for
Go developers building applications that need local language-model inference.
Documented uses
- Integrate llama.cpp-based local inference into a Go application.
- Run a model in the same process as the application rather than through a separate model server.
- Follow a documented setup path from installing tools and libraries to preparing a model and starting a Go program.
Requirements & limits
- The application language is Go, and the integration uses llama.cpp.
- Models run in the application process; the documented approach does not require a separate model server.
- The getting-started path calls for installing both yzma and platform-specific llama.cpp libraries, then obtaining a model.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Prepare to write a Go program with yzma
Set up the documented prerequisites and follow the getting-started path toward a first Go program using a language model.
Before you start
- The yzma command-line tool
- llama.cpp libraries for your platform
- A downloaded model
Steps
Install the yzma command-line tool.
Install the llama.cpp libraries for your platform.
Download a model.
Continue to the first-program guide to write a Go program that uses a language model.
Expected outcome
The documented setup path leads to writing a Go program that uses a language model; the getting-started page does not provide the program's implementation steps in the available content.
Guide limitations
- The getting-started page lists the stages but does not include detailed installation commands or first-program code in the available content.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 5 November 2026