Pipelex: Build & Run AI Methods

A method is a reusable, typed AI procedure declared in a .mthds file and executed by Pipelex. Each step is explicit, each output is structured, and every run is repeatable.

What a Method Looks Like

[pipe.summarize_article]
type    = "PipeLLM"
inputs  = { article = "Text", audience = "Text" }
output  = "Text"
prompt  = "Summarize $article in three bullet points for $audience."

From here, Pipelex handles model routing across 60+ models, structured output parsing, and pipeline orchestration.

Why Methods?

• Declarative — Express business logic in human-readable .mthds files that work across models.
• Typed — Concepts are semantic types: AI understands what you mean, and every input and output connects with purpose.
• Repeatable — Deterministic orchestration that leaves exactly the room you want for AI to express its intelligence.
• Composable — Chain pipes into sequences, nest methods inside methods, and share them with the community.

Capabilities

• Typed Concepts — Semantic types that give meaning to every input and output.
• Pipe Operators — LLM calls, document extraction, image generation, web search, composition, and custom functions.
• Pipeline Orchestration — Sequence, parallel, batch, and conditional controllers.
• 60+ AI Models — OpenAI, Anthropic, Mistral, Google, Deepseek, Hugging Face, and more.
• Validation and Dry Run — Catch errors without spending tokens.
• CLI and Tooling — Init, build, validate, run, and graph visualization.

Get Started

• Build with Claude Code — Describe what you want in natural language.
• MTHDS Language Tutorial — Learn concepts, pipes, sequences, and structured outputs.
• Cookbook Examples — From Hello World to document extraction and image generation.

The MTHDS Ecosystem

• mthds.ai — The MTHDS language specification
• mthds.sh — The Methods Hub for discovering and sharing methods
• GitHub — Pipelex source code

For AI agents: see /llms.txt for a machine-readable index of this documentation.

Redirecting to Pipelex Documentation…