OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Prompt flow

Prompt flow is an open-source Python developer toolkit for creating executable LLM application flows, testing and evaluating their quality, and preparing them for deployment and monitoring. It includes a CLI and a Visual Studio Code flow designer.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Prompt flow — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Prompt flow when an LLM application needs a repeatable development process that covers flow prototyping, debugging, evaluation and preparation for production.

Who it is for

  • Developers prototyping or building applications that use large language models.
  • Teams that want to evaluate LLM flows with datasets and integrate testing into CI/CD.

Documented uses

  • Connect LLMs, prompts, Python code and other tools into executable flows.
  • Trace and debug LLM interactions while iterating on a flow.
  • Evaluate flow quality and performance on larger datasets.
  • Prepare a flow for deployment to a chosen serving platform or integration into an application codebase.
Before you choose

Requirements & limits

The repository is public and MIT-licensed. The README recommends Python 3.9–3.11 for local setup. Its example chat flow requires an OpenAI or Azure OpenAI credential, with an API base also specified for Azure OpenAI.

  • Flows combine LLMs, prompts, Python code and other tools, so the project is aimed at application development rather than a standalone chatbot service.
  • Quality evaluation can use larger datasets and can be integrated into CI/CD.
  • The README describes deployment to a serving platform of the developer’s choice or integration into an app codebase.
  • The Azure AI cloud version is presented as an optional collaboration route; local package-based development is also documented.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Create and interact with a chat flow

Set up the documented chat template and test it interactively using an OpenAI or Azure OpenAI connection.

Before you start

  • Python environment; Python 3.9–3.11 is recommended.
  • Install the promptflow and promptflow-tools packages.
  • An OpenAI API key, or an Azure OpenAI API key and API base.

Steps

  1. Install the packages with pip install promptflow promptflow-tools.

  2. Create the chat template with pf flow init --flow ./my_chatbot --type chat.

  3. Create a connection using the template’s openai.yaml or azure_openai.yaml file and the relevant credentials.

  4. Run pf flow test --flow ./my_chatbot --interactive from the project directory.

Expected outcome

An initialized chat flow that can be tested interactively in the command line.

Guide limitations

  • The example flow specifies a model or deployment name in its flow configuration; the README’s example uses gpt-35-turbo.
  • The README describes Ctrl+C as the way to end the interactive session.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
7 October 2026
Original website title
microsoft/promptflow · Official homepage title · retained in its source language
How we review sourcesSuggest a correction