#3535th of 35 in Agent platforms
promptflow
Python toolkit for building, tracing and evaluating LLM flows from a CLI
- Stars
- 11.3k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Jan 2025
- Language
- Python
11,005 on 31 Mar 202611,253 on 20 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Prompt flow defines LLM apps as executable flows in a flow.dag.yaml that link LLM calls, prompts and Python code. It provides a pf CLI for testing flows, managing OpenAI and Azure OpenAI connections, tracing LLM calls and batch-evaluating against datasets, plus a VS Code flow designer. Microsoft has announced retirement on April 20, 2027 and no longer recommends it for new development.
Who it is for: Teams maintaining Prompt flow apps or planning a migration
Strengths
- Flows are YAML DAGs linking LLM, prompt and Python nodes, runnable with `pf flow test`
- Batch runs and evaluation over datasets can be wired into CI/CD
- Built-in tracing of LLM interactions for debugging flows
- VS Code extension offers a visual flow designer
Weaknesses
- Retirement announced for April 20, 2027; not recommended for new development
- Runtime container images no longer receive security or package updates
- Last release was 2025-01-09; migration to Microsoft Agent Framework is advised
- Telemetry is on by default; opt out with `pf config set telemetry.enabled=false`
What it needs
- no GPU
- Needs Python 3.9-3.11 recommended, OpenAI or Azure OpenAI API key
- Models: OpenAI, Azure OpenAI
Also in Agent platforms
See all 35| Rank | Project | Score |
|---|---|---|
| 1 | LangflowVisual builder for AI agents and workflows, exposed as API or MCP server | 81 out of 100 |
| 2 | daily_stock_analysisDaily LLM stock analysis for A-share, HK, US and other markets, pushed to chat apps | 80 out of 100 |
| 3 | deer-flowAgent harness that runs sub-agents, memory and sandboxed skills for long tasks | 77 out of 100 |
| #4 | vibe-tradingFinance research agent for backtests, market data and broker connectors | 77 out of 100 |
| #5 | MulticaIssue board where AI coding agents take assignments like teammates | 76 out of 100 |