Let's make it real

Getting Started

Everything you need to go from sign-up to a working AI surface. Covers account creation, core concepts, your first flow, and how Runtype's platform keys get you started without needing your own API keys.

Dashboard

A guide to the Runtype home screen — what metrics and activity you'll see at a glance and how to navigate to the areas that matter.

Playground

Test AI models in real time before committing to a flow or agent. Adjust parameters like temperature and max tokens, attach tools, write system prompts, and save working configurations as agents or flows.

Products & Surfaces

Products are how you package and deploy your AI capabilities to the outside world. This section covers creating products, adding flows and agents as capabilities, and deploying them through surfaces — including chat widgets, REST APIs, MCP for AI IDEs, and agent-to-agent (A2A) protocols.

Flows

Flows are multi-step AI workflows built from 30+ step types. Learn how to create, configure, version, and debug flows — from simple prompt chains to complex pipelines with conditionals, API calls, data transforms, and batch processing across thousands of records.

Agents

Agents are AI workflows with autonomous reasoning and tool use. Learn how to create agents, assign tools, and configure models for multi-step task execution.

Records

Records are structured data entries with JSON metadata that provide context for flow and agent execution. Learn how to create, import, filter, and use records across your workflows.

Tools

Tools extend what agents and flows can do. This section covers built-in tools (DALL-E, Exa, Firecrawl), custom JavaScript tools, external HTTP tools, and runtime tools defined inline with API requests.

Evaluations

Compare models and prompt variations side by side. Run evaluations across datasets and analyze the results using aggregate metrics, keyword analysis, and record-level drill-downs to find the best configuration for your use case.

Schedules

Automate flow execution on a recurring basis. Set up cron schedules, pause and resume jobs, and combine scheduling with batch processing for hands-off data pipelines.

Logs

Monitor every execution across your flows, agents, and products. Search and filter by date, status, type, and severity. Use live mode for real-time tailing, export data, and trace failures step by step.

Integrations

Connect Runtype to external services. This section covers AI model providers (OpenAI, Anthropic, Google, XAI), communication tools (Slack, Google Workspace), project management (Linear, GitHub), and data services (Weaviate, Firecrawl, Exa).

Settings

Manage your Runtype account — enable and disable AI models, create and revoke API keys, view billing and usage data, configure third-party integrations, manage team members, and adjust appearance preferences.

Troubleshooting & FAQ

Solutions for common issues including execution failures, authentication problems, and rate limiting. Plus answers to frequently asked questions about the platform.