DSPy Review 2026
Program, don't prompt, your LLMs
What is DSPy?
DSPy is a Python framework for building AI systems by programming LLMs with structured signatures instead of prompts. It provides modules, optimizers, and reusable primitives to produce maintainable, modular, and optimizable programs. It originated at Stanford NLP.
Best for
Mixed
Use cases
- LLM application development
- prompt optimization
- agent building
Key features
signatures
modules
optimizers
multimodal support
ReAct agents
Pricing
Model
free
Price range
Open source, MIT license
Free tier
Yes
Integrations
DSPy integrates with 2 tools including:
OpenAIPython
DSPy alternatives
Other AI Agent Platforms & Frameworks tools worth comparing.
Boundary
The First Language for Building Agents
Haystack
The Open Source AI Framework for Production Ready Agents, RAG & Context Engineering
AutoGen
A programming framework for agentic AI
Agent Skills
A standardized way to give AI agents new capabilities
Lyzr
Take your AI agents to production, faster
Flow AI
Ship production-grade analytical AI agents