stanfordnlp/dspy

📈 Trending★ 34,512PythonAudience · developerComplexity · 3/5ActiveLicenseSetup · moderate

Mindmap

mindmap
  root((DSPy))
    What it does
      Replaces prompts with code
      Auto-optimizes for your task
      Works with any LLM
    Key concepts
      Signatures describe tasks
      Modules chain steps together
      Optimizers refine outputs
    Use cases
      Question-answering systems
      Retrieval-augmented generation
      Multi-step reasoning agents
    Tech stack
      Python
      Multiple LLM backends
      OpenAI, Anthropic, local models

mindmap root((DSPy)) What it does Replaces prompts with code Auto-optimizes for your task Works with any LLM Key concepts Signatures describe tasks Modules chain steps together Optimizers refine outputs Use cases Question-answering systems Retrieval-augmented generation Multi-step reasoning agents Tech stack Python Multiple LLM backends OpenAI, Anthropic, local models

Things people build with this

USE CASE 1

Build question-answering systems that automatically improve their prompts based on example answers.

USE CASE 2

Create retrieval-augmented generation (RAG) pipelines that combine document search with LLM reasoning.

USE CASE 3

Develop multi-step reasoning agents where each step's output quality is automatically optimized.

USE CASE 4

Train classifiers on top of LLMs without manually tuning prompts for each new dataset.

Tech stack

PythonOpenAIAnthropicLLM APIs

Getting it running

Difficulty · moderate Time to first run · 30min

Requires API keys for at least one LLM provider (OpenAI or Anthropic) to run examples.

Use freely for any purpose, including commercial use, as long as you keep the copyright notice.

In plain English

DSPy is a Python framework from Stanford NLP that changes how you build applications powered by large language models (LLMs like GPT-4, Claude, or open-source models). The central idea is to replace hand-written prompts with structured Python code, then let DSPy automatically optimize those prompts or even fine-tune model weights to maximize performance on your specific task. The problem with the conventional approach to LLM development is that prompts are fragile and labor-intensive. A carefully crafted prompt that works well on one task may break with a different model, a slightly different question, or a new version of the same model. Developers end up spending enormous effort tweaking prompt wording rather than building better systems. DSPy treats the prompt as a hyperparameter, something to be automatically tuned, rather than something to hand-craft. The way it works is through "signatures" and "modules." A signature is a short, declarative description of what an LLM call should do (for example, "given a question and context, produce an answer"). Modules are composable building blocks you chain together in Python code to build multi-step pipelines, like a retrieval step followed by a reasoning step followed by an answer generation step. Once your pipeline is written, DSPy's optimizers analyze examples of correct outputs and iteratively refine the prompts and example demonstrations that each module uses, essentially teaching the model what good outputs look like for your specific task. You would use DSPy when building complex LLM systems, question-answering pipelines, retrieval-augmented generation (RAG) systems, multi-step reasoning agents, or classifiers, where prompt quality significantly affects results and you want a principled, automated way to improve them. The tech stack is Python, installable via pip. It supports multiple LLM backends including OpenAI, Anthropic, local models, and others.

Copy-paste prompts

Prompt 1

I want to build a question-answering system using DSPy. Show me how to define a signature and create a module that takes a question and context, then generates an answer.

Prompt 2

How do I use DSPy's optimizers to automatically improve my LLM pipeline's performance on my specific task without manually rewriting prompts?

Prompt 3

Set up a DSPy retrieval-augmented generation system that searches documents and then answers questions based on what it finds.

Prompt 4

I have examples of good outputs for my task. How do I use DSPy to learn from those examples and automatically refine my prompts?

Prompt 5

Show me how to chain multiple DSPy modules together to build a multi-step reasoning pipeline.

Open on GitHub → Explain another repo

Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.