Full Breakdown
Meta Launches Muse Spark 1.1 to Compete in AI Coding Market
7/10/2026, 6:18:07 PM
Meta Unveils Muse Spark 1.1 for Agentic Coding
On Thursday, July 9 2026, Meta released Muse Spark 1.1, the latest iteration of its multimodal AI model aimed at “agentic and coding work.” The update follows an April preview that was limited to select partners via a private API. The new version is offered through a public-preview developer portal, where users can join a waitlist and eventually obtain API keys. Access is currently restricted to Meta-owned properties and is not yet available on third-party marketplaces such as OpenRouter.
Pricing Model and Developer Access
Meta positions the Muse Spark 1.1 API as a cost-competitive alternative to offerings from OpenAI and Anthropic. Each new account receives $20 in free credits; thereafter, Meta charges $1.25 per million input tokens and $4.25 per million output tokens. This pricing is described by Meta as “one of the most attractive pricing levels in the market.” The API is compatible with the OpenAI SDK, allowing developers to point existing stacks at Meta’s model while a Meta-native SDK remains in development.
Official Statements from Meta Executives
AI chief Alexandr Wang framed the launch as a serious business move, noting that token usage is “gigantic and growing very, very quickly” and that capturing a slice of that volume can be “a very meaningful business even for Meta at our scale.” He added that the model is built on Meta’s own computing infrastructure. In a blog post, Meta highlighted that Muse Spark 1.1 “delivers exceptional performance in personal agentic tasks that require planning and orchestration across a range of external apps and services.” CEO Mark Zuckerberg posted on X, calling the model “a strong agentic and coding model at a very low price” and emphasizing its strength in “agentic performance, tool use, and computer use,” while hinting that “more to come soon.”
Competitive Context
Industry analysts note that Meta trails rivals such as OpenAI and Anthropic, which have offered comparable agentic coding models for a longer period. Nonetheless, the competitive pricing may pressure incumbents, whose rates sit near Meta’s $1.25/$4.25 per-million-token structure. The broader AI market remains “as healthy as ever,” with multiple firms releasing new models in rapid succession.
Verbatim Quotes
- “strongest model for agentic and coding work yet,” — Alexandr Wang, AI chief, Meta
- “This is going to be served on top of the computer infrastructure that we've built,” — Alexandr Wang, AI chief, Meta
- “gigantic and growing very, very quickly,” — Alexandr Wang, AI chief, Meta
- “one of the most attractive pricing levels in the market,” — Alexandr Wang, AI chief, Meta
- “Muse Spark 1.1 delivers exceptional performance in personal agentic tasks that require planning and orchestration across a range of external apps and services,” — Meta blog post
- “a strong agentic and coding model at a very low price,” — Mark Zuckerberg, CEO, Meta
