DSPy
DSPy is a free, MIT-licensed, open-source Python framework for building AI programs and systems.
What DSPy is
DSPy is a free, MIT-licensed, open-source Python framework for building AI programs and systems. It represents tasks with signatures that specify typed inputs and outputs, reducing the need to manage prompts directly. Modules can combine step-by-step reasoning, tool use, or ensembles. Optimizers compile programs against examples, a metric, and a scoring function to tune prompts automatically. DSPy uses LiteLLM to normalize model providers and accepts LiteLLM model strings, with OpenAI, Anthropic, Google, and OpenRouter listed among the options. Agent workflows can use dspy.ReAct, dspy.Tool, and dspy.ToolCalls, while remote MCP tools can connect through its tool interface. Images can be signature fields for multimodal tasks. It runs in Python 3.10+, installs with pip or uv, and supports API or self-hosted use. Documentation describes serving REST APIs with FastAPI and deployment with FastAPI or MLflow.
Good
- Free and MIT licensed
- Typed signatures for task inputs and outputs
- Automated prompt optimization
- Multiple provider integrations through LiteLLM
- Supports tool-using agents and multimodal fields
Watch out
- Requires Python 3.10 or newer
- Community language ports lack official support
- Self-hosted or API setup is required
Who it suits
DSPy suits developers building agent, extraction, or multimodal AI systems in Python. It also fits teams wanting automated prompt optimization and connections to multiple model providers.
Verdict
DSPy provides a structured way to compose and optimize AI programs, with broad provider and tool connectivity. Its Python requirement and setup model suit developers rather than users seeking a graphical application.
The facts
Each one read from DSPy's own pages, with the date it was checked.
What it does
- Agent tools
- DSPy supports tool-using agents through `dspy.ReAct` and manual tool handling with `dspy.Tool` and `dspy.ToolCalls`.checked 8 Oct 2026
- API serving
- The deployment guide shows serving a DSPy program as a REST API with FastAPI.checked 9 Oct 2026
- Community integrations
- The community page lists integrations and providers including Weaviate, Qdrant, LangChain, Langfuse, and Weights & Biases Weave.checked 8 Oct 2026
- Community project limits
- The site says community-maintained language ports are not officially supported by the DSPy team and their feature availability and API compatibility may vary.checked 8 Oct 2026
- Community support
- The overview links to DSPy documentation, GitHub, and Discord community resources.checked 9 Oct 2026
- Deployment
- The documentation describes deploying DSPy programs with FastAPI or MLflow.checked 9 Oct 2026
- Install
- DSPy can be installed with `pip install dspy` or `uv add dspy` and works in Python 3.10+ environments.checked 8 Oct 2026
- Installation
- DSPy can be installed with pip or uv and works in Python 3.10+ environments.checked 9 Oct 2026
- Integrations
- DSPy uses LiteLLM to normalize inference providers and says it can connect to nearly any model and provider supported by LiteLLM.checked 8 Oct 2026
- License
- The project is listed as MIT licensed.checked 9 Oct 2026
- Model integrations
- DSPy uses LiteLLM to normalize inference providers and accepts LiteLLM model strings, including options from Anthropic, Google, and OpenRouter.checked 9 Oct 2026
- Modules
- DSPy modules provide different execution strategies, including reasoning, ensembles, tools, and REPL use.checked 9 Oct 2026
- Multimodal
- The overview describes images as a Signature field type for multimodal tasks.checked 8 Oct 2026
- Optimization
- DSPy optimizers compile programs against a metric using examples and a scoring function.checked 9 Oct 2026
- Origin
- DSPy says it started at Stanford NLP and was built in the open beginning in December 2022.checked 8 Oct 2026
- Project origin
- DSPy says it started at Stanford NLP and describes itself as built in the open since December 2022.checked 9 Oct 2026
- Provider examples
- The setup guide names OpenAI, Anthropic, Google, and OpenRouter as provider options.checked 8 Oct 2026
- Signatures
- Signatures define tasks as typed inputs and outputs.checked 9 Oct 2026
- Support
- The site links to a DSPy Discord community and GitHub repository.checked 8 Oct 2026
- Tool integrations
- DSPy can bridge remote MCP tools into its tool interface.checked 9 Oct 2026
- What it does
- DSPy is a Python framework for building AI systems by expressing tasks as structured signatures instead of prompts.checked 9 Oct 2026
- Who it is for
- The overview presents DSPy for building AI programs and systems, including agent, extraction, and multimodal tasks.checked 9 Oct 2026
The company
- Founded
- 2022checked 28 Sep 2026
Alternatives to DSPy
- 1
Vellum
Free planOpen sourceAndroidiOSWebAPI8.7From $30 a month - 2
Scanopy
Free planOpen sourceWebAPILinuxmacOS8.3Free plan - 3
Flowise
Free planWebAPILinuxmacOS7.2From $35 a month - 4
LlamaIndex
Free planWebAPILinuxSelf-hosted7.1From $50 a month - 5
Vercel
Free planWebAPI7.0$20 usage-based - 6
LangChain
Free planAPISelf-hosted6.8$39 usage-based