About LinGoose
LinGoose is a Go framework for building applications on large language models. It is modular, so developers import only the pieces they need, and it abstracts each feature so that a preferred implementation can be chosen or replaced with a custom one. The aim is a complete toolkit for building an AI application in Go from the ground up.
The repository includes a quick start in which a program creates a conversation thread, adds a user message and sends it to an OpenAI model through the framework's LLM package. Related articles by the author cover integrating Anthropic's Claude, building question and answer pipelines, and retrieval augmented generation with Redis. The project has a Discord community and accepts issues and contributions on GitHub.
LinGoose is free and released under the MIT License. It is aimed at Go developers who want to add LLM features to their own software. Model provider usage is billed separately by whichever provider the developer connects.
The README states that LinGoose is no longer under active development. The repository stays available and stable on GitHub, and the author points developers starting new work to Phero, a newer Go framework built for multi agent AI systems.