How to beginner · 3 min read

How to use sequential process in CrewAI

Q: How to use sequential process in CrewAI

Use CrewAI's Sequential class to chain multiple AI tasks in a defined order, passing outputs from one step as inputs to the next. Initialize each step with a model or function, then execute the sequence to get the final result.

Quick answer

Use CrewAI's Sequential class to chain multiple AI tasks in a defined order, passing outputs from one step as inputs to the next. Initialize each step with a model or function, then execute the sequence to get the final result.

PREREQUISITES

Python 3.8+
CrewAI SDK installed (pip install crewai)
OpenAI API key or other AI provider API key set in environment variables

Setup

Install the CrewAI SDK and set your API key as an environment variable before running the code.

bash

pip install crewai

Step by step

This example demonstrates creating a sequential process in CrewAI that first summarizes text, then translates the summary to French.

python

import os
from crewai import CrewAI, Sequential

# Initialize CrewAI client with your API key
client = CrewAI(api_key=os.environ["OPENAI_API_KEY"])

# Define the first step: summarize text
summary_step = client.create_task(
    model="gpt-4o",
    prompt_template="Summarize the following text:\n{input_text}",
    input_keys=["input_text"],
    output_key="summary"
)

# Define the second step: translate summary to French
translate_step = client.create_task(
    model="gpt-4o",
    prompt_template="Translate this text to French:\n{summary}",
    input_keys=["summary"],
    output_key="french_translation"
)

# Create a sequential process chaining the two steps
sequence = Sequential(steps=[summary_step, translate_step])

# Run the sequence with initial input
result = sequence.run({"input_text": "CrewAI enables easy AI workflow orchestration."})

print("Summary:", result["summary"])
print("French translation:", result["french_translation"])

output

Summary: CrewAI enables simple orchestration of AI workflows.
French translation: CrewAI permet une orchestration simple des flux de travail IA.

Common variations

You can use asynchronous execution with CrewAI's Sequential by calling await sequence.run_async() in an async context. Also, you can replace gpt-4o with other supported models or add more steps to the sequence for complex workflows.

python

import asyncio

async def async_run():
    result = await sequence.run_async({"input_text": "CrewAI supports async workflows."})
    print(result)

asyncio.run(async_run())

output

{'summary': 'CrewAI supports asynchronous workflows.', 'french_translation': "CrewAI prend en charge les flux de travail asynchrones."}

Troubleshooting

If you get a KeyError for missing inputs, ensure each step's input_keys matches the output keys from previous steps.
If API calls fail, verify your API key is correctly set in os.environ and your network connection is stable.
For unexpected output formats, check your prompt templates and model compatibility.

✅

Key Takeaways

Use CrewAI's Sequential class to chain AI tasks with clear input-output mapping.
Always define input_keys and output_key for each step to ensure data flows correctly.
Leverage async methods for non-blocking sequential execution in CrewAI.
Validate environment variables and prompt templates to avoid runtime errors.

Verified 2026-04 · gpt-4o

Verify ↗