Instructions to use KOHENOOR-AI/kai-lite1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use KOHENOOR-AI/kai-lite1 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf KOHENOOR-AI/kai-lite1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf KOHENOOR-AI/kai-lite1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf KOHENOOR-AI/kai-lite1:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf KOHENOOR-AI/kai-lite1:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Use Docker

docker model run hf.co/KOHENOOR-AI/kai-lite1:Q4_K_M

LM Studio
Jan

vLLM

How to use KOHENOOR-AI/kai-lite1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KOHENOOR-AI/kai-lite1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KOHENOOR-AI/kai-lite1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KOHENOOR-AI/kai-lite1:Q4_K_M

Ollama
How to use KOHENOOR-AI/kai-lite1 with Ollama:
```
ollama run hf.co/KOHENOOR-AI/kai-lite1:Q4_K_M
```
Unsloth Desktop

How to use KOHENOOR-AI/kai-lite1 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "KOHENOOR-AI/kai-lite1:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use KOHENOOR-AI/kai-lite1 with Docker Model Runner:
```
docker model run hf.co/KOHENOOR-AI/kai-lite1:Q4_K_M
```

Lemonade

How to use KOHENOOR-AI/kai-lite1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull KOHENOOR-AI/kai-lite1:Q4_K_M

Run and chat with the model

lemonade run user.kai-lite1-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use KOHENOOR-AI/kai-lite1 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default KOHENOOR-AI/kai-lite1:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use KOHENOOR-AI/kai-lite1 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KOHENOOR-AI/kai-lite1:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "KOHENOOR-AI/kai-lite1:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

KAIPAL Lite

KAIPAL Lite is good for one thing: day-to-day business and finance advice on your own machine. Pricing, cash, customers, suppliers, everyday business writing, and the sums behind them, worked out line by line, for business owners, professionals and students, offline and private.

KAI Pocket Assistant Lite, by Kohenoor Technologies. A small model, about 2 billion active parameters, for day-to-day business and finance advisories: markets, pricing, cash, customers, everyday business work, and the sums behind them. It runs on your own machine, offline, in about 4 GB of memory.

Run it with Ollama

The quickest way, with Ollama 0.30.9 or newer installed and nothing to download by hand:

ollama run kohenoor/kai-lite1 --think=false

Or from this repository: download the two files into one folder and run:

ollama create kaipal -f Modelfile
ollama run kaipal --think=false

Keep --think=false: KAIPAL Lite answers directly, and that is how it is meant to be used. Ask as a business owner, a professional or a student would. Give your own figures and it works the sum out line by line before it concludes.

What it is for

Advice on markets, pricing, positioning, customers and competition.
Starting, running and growing a small business: cash, costs, hiring, growth timing, operations.
Everyday business work: messages to customers and suppliers, replies to complaints, plans, quotations.
Business and finance explained: margins, break-even, cash flow, credit notes, letters of credit and the like.
Personal development at work: discipline, focus, communication, negotiation, careers.

What it does not do

It makes no buy, sell or hold calls and predicts no prices. It does not recommend borrowing at interest. It gives no legal, tax or medical conclusions, and it does not help anyone deceive. For those it points to KAI Premium at www.kohenoor.net or to a qualified professional. Outside business and personal development it declines in a line.

Limits of this release

Text only. Paste the text of a document; this file does not read pictures or recordings.
Arithmetic is worked line by line and is right far more often than not, but check the conclusion against the figures before you act on it.
Run settings in the Modelfile: temperature 0.15, a 16K context.

Terms

This release is distributed under the Apache License, Version 2.0: the full text is in LICENSE, and NOTICE states what the files are and that they have been modified. Kohenoor Technologies' own terms for KAI products apply in addition. KAI, KAIPAL and KAI Pocket Assistant are names of Kohenoor Technologies.

Checksum of the model file (SHA-256) is in CHECKSUM.txt. The guidelines for using and testing it, and the place to report what you find, are on GitHub at https://github.com/Kohenoor-Technologies/kai-lite1. The model is also on the Ollama registry at https://ollama.com/kohenoor/kai-lite1.

Downloads last month: -

GGUF

Model size

5B params

Architecture

gemma4

Hardware compatibility

4-bit