Sundar Pichai pens welcome note to Mark Zuckerberg, says 'excited to serve Google Cloud customers'

Google CEO Sundar Pichai on Tuesday welcomed Meta CEO Mark Zuckerberg's announcement that the company's Muse Ember, Muse Spark 1.3 and Muse Wonder AI models are now available through Google Cloud's Model Garden managed APIs.

Zuckerberg recently announced the availability of the three models in a post on X. Pichai responded by welcoming the models to Google Cloud's platform and highlighting the collaboration between the two companies.

Also read: Adobe, Canva, Figma tools in Claude: Anthropic lets users edit photos, PDFs and designs without switching apps


“Muse Ember, Muse Spark 1.3, and Muse Wonder are now available on Google Cloud,” Zuckerberg wrote.

“Congrats and welcome to the model garden, excited to serve our Google Cloud customers!” Pichai replied to the Meta CEO.

The models are multimodal and feature context windows of up to 1 million tokens. Muse Ember is Meta's smallest and most affordable model, offering intelligence comparable to models such as Claude Haiku at around one-third of the cost per task. It is designed for sub-agent tasks and repetitive workloads.

What is Muse Wonder?

Muse Wonder is positioned as Meta's most capable frontier model, designed for complex software engineering tasks. It aims to balance speed, reliability and the ability to handle complex, extended workflows.

The model supports native computer use and long-context understanding, allowing it to execute multi-step tasks across browsers, desktop and mobile applications.

Its capabilities include coding, orchestration and multimodal reasoning, along with sub-agent delegation, parallel tool calling and agentic memory.

Key use cases include software engineering, updating multiple files across repositories and automating end-to-end workflows. It can also operate everyday applications, assist with complex legal and professional work, and analyse large codebases.

Also read: Anthropic launches Claude Whiteboards, Sheets and Data to expand AI-powered workplace tools

What is Muse Spark 1.3?

Muse Spark 1.3 is an agentic model designed to handle extended workflows, balancing intelligence, speed and scalable performance.

It supports native tool calling and Model Context Protocol (MCP) server integration, allowing developers to build AI agents capable of completing multi-step tasks across professional and personal workflows.

The model offers coding, reasoning and autonomous execution capabilities, with specialised versions designed for high-volume production environments on the Gemini platform.

Its primary applications include workflow automation, developing tool-enabled applications and executing long-running agentic tasks. It can also power personal and enterprise assistants, manage complex API and MCP pipelines, support software development and handle dynamic real-world tasks at scale.

What is Muse Ember?

Muse Ember is a lightweight, multimodal execution model designed for routine, repetitive tasks, with a focus on balancing capabilities, throughput and cost efficiency.

It supports text, image, audio and video processing, allowing developers to match tasks with the appropriate model capabilities across enterprise workflows.

The model offers coding, tool-calling and document-analysis capabilities. Its lightweight architecture is designed to complement Muse Wonder and Muse Spark on Google Cloud.

Its key applications include everyday coding tasks, automated tool execution and large-scale document analysis. It can also support high-volume task routing, act as the execution layer for complex multi-agent systems, analyse multimodal content and manage long-running workflows at scale.