gitwtfhub

wtf is pi-meta-ai?

seemethere/pi-meta-ai — explained in plain English

Analysis updated 2026-05-18

0TypeScriptAudience · developerComplexity · 2/5LicenseSetup · moderate

TL;DR

An extension that adds Meta's Muse Spark 1.1 AI model as an option inside the pi coding agent, using an API key.

Mindmap

mindmap
  root((pi-meta-ai))
    What it does
      Adds Meta model to pi
      Registers API provider
      Enables agent features
    Tech stack
      TypeScript
      Node.js
      Meta Model API
    Use cases
      Use Muse Spark in pi
      Try large context window
      Test image input coding
    Audience
      Developers using pi
      AI tool tinkerers
    Setup
      Get API key
      Install extension
      Select model in pi

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

Why would anyone build with this?

REASON 1

Add Meta's Muse Spark 1.1 model as a selectable option inside the pi coding agent.

REASON 2

Experiment with a 1 million token context window for large codebases or documents.

REASON 3

Try agentic features like tool calling and structured output with Meta's model for free during its preview period.

What's in the stack?

TypeScriptNode.jsMeta Model API

How it stacks up

seemethere/pi-meta-ai0xkinno/astraea0xkinno/halcyon
Stars000
LanguageTypeScriptTypeScriptTypeScript
Setup difficultymoderatehardhard
Complexity2/54/54/5
Audiencedeveloperdeveloperdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you spin it up?

Difficulty · moderate Time to first run · 30min

Requires an API key from Meta's developer portal and the pi coding agent already installed.

Use freely for any purpose, including commercial use, as long as you keep the copyright notice.

Wtf does this do

pi-meta-ai is a small extension that connects the pi coding agent to Meta's Model API, giving it access to a model called Muse Spark 1.1. If you already use pi as your AI coding assistant, this extension lets you add Meta's model as another option alongside whatever you use today, using an API key from Meta's developer site. Once installed, the extension registers Meta's API as a provider inside pi and supports the same kind of agentic features pi normally offers: calling tools, running several tools at once, streaming responses as they are generated, adjustable reasoning effort, structured JSON output, and even image input in formats like PNG or JPEG. It also supports a very large context window of up to 1 million tokens, which means it can take in a large amount of text or code at once. Setup involves installing the pi coding agent through npm, then either installing this extension from source or as a package, getting an API key from Meta's developer portal, and providing that key to pi either through a login command inside the tool or by setting an environment variable before starting pi. After that, you select the Meta model from within pi and start chatting normally, with pi's usual tools like reading files, running commands, and editing code all working as before. The README notes that Meta's Model API is currently in a free preview, so cost is shown as zero for now and will be updated if pricing changes. The project includes basic troubleshooting steps for authentication issues and a short development guide for anyone who wants to modify the extension itself. It is released under the MIT license.

Yoink these prompts

Prompt 1
Walk me through installing pi-meta-ai and connecting it to my Meta API key.
Prompt 2
Explain the difference between authenticating pi-meta-ai with /login versus an environment variable.
Prompt 3
Help me troubleshoot why the Meta model is not showing up in pi's /model list.
Prompt 4
Show me how to configure models.json manually instead of using the extension.

Frequently asked questions

wtf is pi-meta-ai?

An extension that adds Meta's Muse Spark 1.1 AI model as an option inside the pi coding agent, using an API key.

What language is pi-meta-ai written in?

Mainly TypeScript. The stack also includes TypeScript, Node.js, Meta Model API.

What license does pi-meta-ai use?

Use freely for any purpose, including commercial use, as long as you keep the copyright notice.

How hard is pi-meta-ai to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is pi-meta-ai for?

Mainly developer.

View the repo → Decode another repo

This repo across BitVibe Labs

Don't trust strangers blindly. Verify against the repo.