To build a team, create member agents and compose them under a Team with an execution mode.
from definable.agent import Agent
from definable.agent.team import Team, TeamMode
researcher = Agent(model="gpt-4o", instructions="You are a research specialist. Find facts and data.")
writer = Agent(model="gpt-4o", instructions="You are a technical writer. Write clear, accurate content.")
editor = Agent(model="gpt-4o", instructions="You are an editor. Polish prose and fix errors.")
team = Team(
name="content-team",
model="gpt-4o",
members=[researcher, writer, editor],
mode=TeamMode.coordinate,
instructions="Produce well-researched, polished technical content.",
)
result = await team.arun("Write about the state of AI agents in 2026.")
print(result.content)
Run Your Team
Set up your environment
python3 -m venv .venv
source .venv/bin/activate
Export your API key
export OPENAI_API_KEY=sk-***
Constructor Reference
Human-readable team name. Auto-generated if not set.
Model for the leader agent. Accepts string shorthand ("gpt-4o") or a Model instance.
Agents or nested teams the leader can delegate to.
mode
TeamMode
default:"coordinate"
Execution mode: coordinate, route, collaborate, or tasks.
Instructions for the leader agent.
Maximum delegation rounds (relevant for tasks mode).
share_member_interactions
When true, member responses are shared with subsequent delegations.
Additional tools for the leader (on top of auto-injected delegation tools).
Pydantic model for structured output from the leader.
Enable debug logging for team operations.
Next Steps
| Task | Guide |
|---|
| Execute and handle team responses | Running teams |
| Understand delegation patterns | Delegation |
| Orchestrate with step-based workflows | Workflows |
