🌱 Langapp: LLM based AI-chat/assistants testing with Langchain on CLI

👉 Requirements: Python, and any API keys and functions you want; I am currently using only the openai_api_key.

🔹 Create and activate a virtual environment in the preferred project directory

python -m venv venv

On Windows

venv/Scripts/activate

On Linux or Mac

source venv/bin/activate

🔹 Install dependencies

pip install -r requirements.txt

Install/build langapp

pip install .

---

🔹 How to use

Default model gpt-4o-mini from OpenAI.

langapp

You can select other models, such as llama3.1 or deepseek-r1 from Ollama after installing and setting it up.

langapp --m llama3.1

You can select chat role as well. Default is: short and concise

langapp --r "correct english"

Other roles: correct german upgrade english, remaster the conversation translate to english/german/hungarian or extend with what you need in prompts.py and helpers.py.

End conversation: empty prompt / press Enter

---

🔹 Ollama

Download and install from Ollama's website to use it locally.

test a model

ollama run <model_name>:<param>b

In the case of using an open-source model from Ollama, you need to run the command before using Langapp.

---

🔹 Environment variables

#.env

OPENAI_API_KEY

---

🔹 Using scripts

shell_scripts/runapp.bat on Windows

shell_scripts/runapp.sh on Linux

Don't forget to add your project path to the script files!

---

👉 More in .app_info