Langflow is an intuitive interface allows for easy manipulation of AI building blocks. With Langflow developers can prototype and turn their ideas into powerful production solutions in no time.
Requires application load balancer
Due to the size of payloads returned by Langflow it is not suitable for running behind API Gateway. Thus it requires an Application Load Balancer which incurs an additional fee of around $16/month per environment.
Features
- Fine-tuning: Utilize the full potential of LLMs by fine-tuning them from spreadsheets
- Group components: High level components emerge naturally from your AI building blocks
- Code your own components
- Integrations: Integrate seamlessly with tools and stacks your team loves without writing a line of code
Variables
LANGFLOW_SUPERUSER
: The username of your admin accountLANGFLOW_SUPERUSER_PASSWORD
: The password for your admin accountLANGFLOW_DATABASE_URL
: Optional To use Postgres instead of SQLite add a PostgreSQL connection URI for the database you want to use as a backend and apply Umami migrations to. Check out the FlexStack Databases guide for integration recommendations. Below is a screenshot of a copyable connection URI in the Neon dashboard.
Getting started
Once your app is deployed, open your app's URL in your browser. Then login using the values from the LANGFLOW_SUPERUSER
and LANGFLOW_SUPERUSER_PASSWORD
variables.
Additional configuration
To set up API keys for services like Anthropic, Cohere, and OpenAI you can set Global Variables in your Secrets and Variables or add them in your Langflow dashboard under Global Variables settings.