Executable AI Methods

Declare multi-step AI methods in typed .mthds files — deterministic orchestration, structured outputs, repeatable results.

Demo (2min) - Docs - Hub - Discord

[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.

Open Claude Code and paste this. Claude does the rest.

Install mthds: run bash -c 'npm install -g mthds && mthds-agent bootstrap && claude plugin marketplace add mthds-ai/mthds-plugins && claude plugin install mthds@mthds-plugins'. Then tell me to /reload-plugins.

Then build your first method:

/mthds-build A method to summarize articles with key takeaways for different audiences

Run it:

/mthds-run

Open Codex and paste this. Codex does the rest.

Install mthds: run bash -c 'npm install -g mthds && mthds-agent bootstrap && mthds-agent codex apply-config && codex plugin marketplace add mthds-ai/mthds-plugins'. Then tell me to restart Codex and run /plugins to install mthds.

Requires Codex 0.130.0+ (npm install -g @openai/codex@latest if needed). See the mthds-plugins README for what mthds-agent codex apply-config does and manual install steps.

uv tool install pipelex
pipelex init

1. Install the VS Code extension for .mthds syntax highlighting
2. Browse methods on the MTHDS Hub for inspiration
3. Author your own .mthds methods
4. Validate with pipelex validate bundle your_method.mthds
5. Run with pipelex run bundle your_method.mthds

Two ready-to-fork templates depending on how you want to run methods:

• pipelex-starter-python — Python project embedding the pipelex runtime directly. Best when you want to run methods in-process from a Python service, script, or notebook. Click Use this template on GitHub.
• pipelex-starter-js — Next.js 16 + TypeScript app that calls a Pipelex API server via the mthds SDK. Best when you want a TypeScript frontend/backend that talks to a remote (hosted or self-hosted) Pipelex runner. Ships with three demo pipelines (text entities, PDF summary, image generation).

• Pipelex Gateway (Recommended) — Free credits, single API key for LLMs, OCR / document extraction, and image generation across all major providers.
• Bring Your Own Keys — Use existing API keys from OpenAI, Anthropic, Google, Mistral, etc. See Configure AI Providers.
• Local AI — Ollama, vLLM, LM Studio, or llama.cpp — no API keys required. See Configure AI Providers.

Ready-to-run methods in the Cookbook: classification, extraction, analysis, generation, and more.

• Discord — Get help, share methods, meet the team
• Documentation — Guides and reference
• GitHub Issues — Report bugs and request features
• security@pipelex.com — Security and privacy concerns

All repositories are MIT licensed unless otherwise specified. See individual LICENSE files for details.

"Pipelex" is a trademark of Evotis S.A.S.