Lesson 1 — Quick Start: Your First Chat in 5 Lines

Prerequisite: Python 3.10+, an OpenAI or Anthropic API key.

Installation

pip install lllm-core

Set your API key in the environment:

export OPENAI_API_KEY=sk-...
# or
export ANTHROPIC_API_KEY=sk-ant-...

One-Line Chat

from lllm import Tactic

response = Tactic.quick("What is the capital of France?")
print(response.content)   # "Paris"

Tactic.quick() is the zero-config entry point. It creates a temporary agent, sends the message, and returns the response as a Message object. The .content attribute holds the plain text reply.

Choosing a Model

response = Tactic.quick(
    "Explain quantum entanglement in one sentence.",
    model="claude-opus-4-6",   # any LiteLLM model string
)
print(response.content)

LLLM uses LiteLLM under the hood, so you can pass any model identifier from that catalog — GPT-4o, Claude, Mistral, Gemini, etc.

Custom System Prompt

response = Tactic.quick(
    "What is 2 + 2?",
    system_prompt="You are a grumpy math teacher who always sighs before answering.",
)
print(response.content)

The system_prompt string is converted into a Prompt object automatically. You will learn more about Prompt in Lesson 3.

Getting the Agent Back

Sometimes you want to reuse the agent for follow-up messages:

response, agent = Tactic.quick(
    "What is the capital of France?",
    return_agent=True,
)
print(response.content)
print(agent.name)   # "assistant"

Or get an agent without sending any query first:

agent = Tactic.quick(system_prompt="You are a helpful assistant.", model="gpt-4o")
# … set up the dialog later
agent.open("chat")
agent.receive("What is the capital of France?")
print(agent.respond().content)

This pattern — open → receive → respond — is the foundation for everything in LLLM. Lesson 2 covers it in depth.

What `Tactic.quick` Does Internally

Tactic.quick(query, system_prompt, model)
  ├── builds a Prompt from system_prompt string
  ├── builds a LiteLLM invoker
  ├── creates an Agent(name="assistant", model=model, ...)
  ├── calls agent.open("chat")
  ├── calls agent.receive(query)
  └── returns agent.respond()

There is no config file, no folder structure, and no subclassing needed for this path. All of that is optional and covered in later lessons.

Supported Providers

Provider	Model example	Environment variable
OpenAI	`gpt-4o`	`OPENAI_API_KEY`
Anthropic	`claude-opus-4-6`	`ANTHROPIC_API_KEY`
Mistral	`mistral/mistral-large`	`MISTRAL_API_KEY`
Any other	see LiteLLM docs	provider-specific

Summary

pip install lllm-core + set an API key is all the setup needed.
Tactic.quick(query) gives you a one-liner LLM call.
The response is a Message object; .content is the text.
Pass return_agent=True to get the Agent back for follow-up turns.

Next: Lesson 2 — Agents and Dialogs