# DSPy

Python framework that expresses language-model tasks as modules and optimizes them against metrics.

- Company: Stanford NLP
- Category: Platforms & Infrastructure
- Business model: Open (open weights/source)
- Availability: unknown
- Launched: Not recorded
- Tags: framework, developer-tools
- Website: https://dspy.ai/
- Documentation: Not recorded
- Pricing page: Not recorded
- License: Not recorded
- Canonical URL: https://globalaiproductindex.com/products/dspy/
- Record updated: 2026-11-03

Typical uses include Development of tool-using agents; Construction of retrieval-augmented applications. Documented capabilities: Structured task signatures instead of hand-written prompts; Composable prediction, reasoning and retrieval modules; Optimizers that improve prompts using examples and metrics.

## Key features

- Structured task signatures instead of hand-written prompts
- Composable prediction, reasoning and retrieval modules
- Optimizers that improve prompts using examples and metrics

## Use cases

- Development of tool-using agents
- Construction of retrieval-augmented applications

## Pricing

The open-source software can be run on your own infrastructure. Compute, storage and any external model API usage are separate costs; a managed service is not included.

## FAQ

**Where does DSPy fit in a workflow?**

Python framework that expresses language-model tasks as modules and optimizes them against metrics. Structured task signatures instead of hand-written prompts.

**What costs and limits should I compare before choosing DSPy?**

The open-source software can be run on your own infrastructure. Compute, storage and any external model API usage are separate costs; a managed service is not included.

## Sources

- [DSPy — official product](https://dspy.ai/)
- [DSPy — official documentation and terms](https://github.com/stanfordnlp/dspy)

---
Data from the [Global AI Product Index](https://globalaiproductindex.com/), licensed CC BY 4.0.