Enterprise access
Agent Serve is available to teams by invitation. Tell us what you want to build and we will set up your workspace with you.
Agent Serve · AI integration
Run predictions using self-hosted DSPy programs
DSPy is a AI integration for Agent Serve, the AI workspace where teams build and deploy AI agents. Agent Serve's DSPy integration provides 3 DSPy tools that AI agents can use inside Agent Serve's visual workflow builder. DSPy connects with an API key. Explore deployment options with the Agent Serve team.
Last updated
Integrate with your self-hosted DSPy programs for LLM-powered predictions. Supports Predict, Chain of Thought, and ReAct agents. DSPy is the framework for programming—not prompting—language models.
Agent Serve accounts are created by invitation. Request access for your team, or accept an invitation from your workspace admin, then sign in.
Open your workspace, drag a DSPy block onto the workflow builder, and paste in your DSPy API key.
Pick the tool you need, wire in an AI agent for reasoning or data transformation, and run. Your DSPy automation is live.
3 DSPy tools available in Agent Serve.
Run a prediction using a self-hosted DSPy program endpoint
Run a Chain of Thought prediction using a self-hosted DSPy ChainOfThought program endpoint
Run a ReAct agent using a self-hosted DSPy ReAct program endpoint for multi-step reasoning and action
Ready-to-use templates featuring DSPy. Click any to build it instantly.
Agent Serve's DSPy integration adds 3 DSPy tools to the AI agents you build in Agent Serve's visual workflow builder — you build it all visually. Run predictions using self-hosted DSPy programs. Teams often pair DSPy with Cursor and Mintlify in the same agent.
Agent Serve is available to teams by invitation. Request access and the team will map out the DSPy agents, connections, and team access your process needs.