Simon Willison Releases llm-meta-ai 0.2: Open-Source CLI Plugin for Meta AI API Models

On October 12, 2026, open-source developer and data journalist Simon Willison officially rolled out version 0.2 of llm-meta-ai, an integration plugin designed for his Python-based llm command-line ecosystem. The update was simultaneously published to GitHub releases (simonw/llm-meta-ai) and the PyPI package repository under version 0.2, ensuring release consistency across package management environments.

Image source: Simon Willison / GitHub

This 0.2 release serves as a third-party developer utility allowing engineers to test and evaluate Meta AI API models directly within standard terminal shell workflows. It provides an accessible command-line path for probing input-output dynamics without building custom API harnesses from scratch.

Meta AI API Integration and llm-meta-ai 0.2 Release Highlights

The llm-meta-ai project extends Simon Willison's broader llm CLI utility, which standardizes model interactions across diverse cloud and local inference backends.

The central advancement in version 0.2 is structured connectivity to the Meta AI API infrastructure. Developers can now incorporate Meta AI endpoints directly into command-line pipelines and shell scripting tasks.

• Synchronized Distribution: Version 0.2 is registered identically on GitHub releases and PyPI, enabling reliable dependency pinning and package resolution.
• Terminal-Centric Pipeline: Users can send prompts and consume model responses through standard input/output streams, bridging Meta AI capabilities with existing Unix tools and automated test scripts.

Muse Wonder Model Support and CLI Testing Workflows

Technically, the standout capability introduced in this release is support for Meta AI's Muse Wonder model family.

According to release documentation, version 0.2 implements interfaces for both the standard Muse Wonder model and the Muse Wonder Full variant. The package also bundles example test configurations covering image generation across distinct inference parameters, giving developers direct insight into model behavior.

• Model Targeting: The plugin supports explicit selection between Muse Wonder and Muse Wonder Full model identifiers for inference queries.
• Image Generation Test Cases: Included test examples demonstrate prompt handling and image generation workflows under varied inference configurations.

Third-Party Open-Source Boundaries and Platform Prerequisites

Engineering teams evaluating this plugin must remain mindful of its architectural boundaries and operational prerequisites.

This release represents an independent third-party open-source utility, not an official model launch or public API announcement by Meta. Teams considering adoption should account for several key factors:

• Platform Policy Dependencies: Actual API access credentials, rate limits, request quotas, and binding terms of service remain entirely governed by Meta's native platform policies.
• Third-Party Maintenance Model: Because llm-meta-ai is maintained as a personal open-source project by Simon Willison, it carries no formal service-level agreement (SLA) or commercial warranty from Meta, and upstream API changes may require interface updates.
• Verification of Availability: Developers must independently confirm model availability, pricing schedules, and regional access tiers through Meta's official developer documentation prior to designing internal tools around the API.

Sources

• GitHub Release: simonw/llm-meta-ai 0.2 Release
• Simon Willison's Weblog: llm-meta-ai Release Notes
• Let's Data Science: Simon Willison releases llm-meta-ai 0.2 for Meta AI API models