Launched January 3, 2023
AI Config

AI Config

Build Production-Grade AI Apps

Free4,954 impressions#180 of its week

AIConfig is an open-source framework that automates the setup and implementation of Generative AI models. Instead of calling APIs directly, we handle that for you through prompt, model and parameter settings in a JSON config.

AIConfig is an open-source framework that manages generative AI prompts, models, and parameters via JSON configs.

for
Developers building production-grade generative AI applications.
pricing
open source
license
MIT
lastmile-ai/aiconfig 1.1k 92Pythonupdated 8 months ago
works withPythonNode.jsVS CodeOpenAI APILLaMAHugging Face

Key features

6 features of AI Config
  • Prompts as Configs — Store prompts and model settings in a standardized JSON format for version control.
  • AIConfig Editor — VS Code extension to visually create, edit, and run prompt chains with auto-save.
  • Model-agnostic SDKs — Python and Node SDKs let you run configs with any supported generative model.
  • Extensibility — Add support for new models, callbacks, and custom metadata via extensions.
  • Collaborative Development — Share and iterate on aiconfig files across teams for coordinated AI development.
  • Supported Multimodal Models — Works with OpenAI, Gemini, LLaMA, PaLM, Hugging Face and more out of the box.

Use cases

  • Build a RAG pipeline with chained prompts
  • Create a function-calling chatbot using OpenAI models
  • Develop multi-LLM consistency checks across text, image, and audio

AI Config FAQ

How do I install AIConfig?+

Use pip3 install python-aiconfig for Python or npm install aiconfig for Node.

Do I need an OpenAI API key?+

Yes, set the OPENAI_API_KEY environment variable to use OpenAI models.

Can I edit configs without VS Code?+

Yes, run aiconfig edit to open the web-based editor in your browser.

Is AIConfig limited to OpenAI models?+

No, it supports multiple providers and can be extended to custom models via ModelParser.

How are changes saved?+

The editor auto-saves every 15 seconds and you can manually save; outputs can be serialized back to the JSON file.

Summarized by DevHunt from github.com · Oct 4, 2026. Details may change; check the official site.