In this article4 sections
Meta officially entered the increasingly competitive AI coding market on Thursday with the launch of Muse Spark 1.1, a multimodal AI model designed for agentic coding tasks. The company is charging $1.25 per million input tokens and $4.25 per million output tokens, positioning itself directly against established rivals OpenAI and Anthropic.
The new model, an update to the version first announced in April, is designed to handle multi-step reasoning, manage complex digital workflows, and deploy new features within enterprise systems. Meta is entering a space where Anthropic and OpenAI have already established a strong presence, but the company is betting that competitive pricing and a focus on large-scale automation will attract enterprise customers.
Also read: OpenAI unveils Private Safety Processing to counter Anthropic's data retention policy
Competitive pricing and enterprise focus
Meta’s pricing strategy puts Muse Spark 1.1 in line with, though slightly above, comparable offerings from competitors. According to Reuters, the model costs $1.25 per million input tokens and $4.25 per million output tokens, placing it near Anthropic’s Claude Haiku 4.5 and OpenAI’s GPT-5.6 Luna.
The company’s pitch to enterprise users focuses on Spark’s ability to handle large agentic workloads, fix bugs, and assist with large code migrations — the kind of automation that businesses are increasingly seeking from AI providers. In a blog post, Meta stated 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.”
Also read: Anthropic locks in $45B Nscale compute deal to fuel AI race against OpenAI
Zuckerberg’s return to X
The release of Muse Spark 1.1 was significant enough to prompt Meta CEO Mark Zuckerberg to post on the social media platform X for the first time in three years. His last post was in July 2023, around the time the platform rebranded from Twitter to X. In his post, Zuckerberg described Spark as “a strong agentic and coding model at a very low price,” noting that it is “strongest at agentic performance, tool use, and computer use.” He also hinted at future releases, stating there is “more to come soon.”
A busy week for AI announcements
Meta’s Muse Spark 1.1 launch is part of a broader wave of AI announcements this week. The company also unveiled a new AI image generation model, Muse Image, on Tuesday. Other notable releases include a new version of Grok from SpaceXAI and OpenAI’s GPT-5.6 family of models, which also launched on Thursday. The flurry of activity underscores the intense competition in the AI industry, where companies are racing to differentiate their offerings through pricing, performance, and specialization.
Frequently Asked Questions
What makes Muse Spark 1.1 different from other AI coding models?
Muse Spark 1.1 is designed specifically for agentic coding tasks, meaning it can plan, reason, and execute multi-step workflows autonomously. Meta emphasizes its ability to handle large-scale code migrations and bug fixes, targeting enterprise users.
How does Meta’s pricing compare to OpenAI and Anthropic?
Meta’s pricing for Muse Spark 1.1 is $1.25 per million input tokens and $4.25 per million output tokens, which is competitive and slightly above Anthropic’s Claude Haiku 4.5 and OpenAI’s GPT-5.6 Luna.
Why did Mark Zuckerberg post on X for the first time in three years?
Zuckerberg posted to promote the launch of Muse Spark 1.1, calling it a strong and affordable model. His return to the platform after nearly three years signals the strategic importance of this release for Meta.