Meta made a strategic pivot by releasing a commercial version of its language model, Muse Spark 1.1. Instead of the traditional focus on open source and free technology, the company went all-in, setting aggressive API prices that are significantly lower than competitors' rates from Anthropic and OpenAI. This is not just an update—it's a direct attack on the leaders of the AI coding and autonomous agent market.
Previously, Meta exclusively developed free, open-source models, but the strategy has now changed dramatically. Head of AI Alexander Wang directly stated that Muse Spark's pricing aims to displace key competitors. And the numbers confirm this: processing 1 million input tokens costs just $1.25—37% cheaper than Anthropic's Sonnet 5 starting rate ($2) and 75% lower than top-tier Opus 4.8 and GPT-5.5 (both at $5).
Price Gap: Leaderboard Table
The comparison for output tokens is even more impressive: $4.25 per 1 million from Meta versus $10 from Sonnet 5, $25 from Opus 4.8, and $30 from GPT-5.5. This translates to savings of 58% to 86% for developers. Even after the promotional period for Sonnet 5 ends on August 31, when its rate returns to the standard $3/$15, Meta's offer will remain significantly more advantageous.
| AI Model | Input Tokens (per 1M) | Output Tokens (per 1M) |
| Meta Muse Spark 1.1 | $1.25 | $4.25 |
| Anthropic Sonnet 5 | $2.00 | $10.00 |
| Anthropic Opus 4.8 | $5.00 | $25.00 |
| OpenAI GPT-5.5 | $5.00 | $30.00 |
According to feedback from programmers who have tested Muse Spark 1.1, the model demonstrates performance comparable to GPT-5.5 and Claude Opus 4.8 in complex autonomous execution tasks. However, Meta has not yet provided independent benchmarks, leaving room for skepticism. Access to the platform is strictly limited—only specialists from the US after application approval on a waiting list.
My expert opinion: This price attack is a classic "scorched earth" move from Meta. By lowering the entry barrier, the company aims to lure a significant portion of the IT community while competitors try to justify their premium rates. However, quality is the key factor. If Muse Spark truly proves to be on par with top-tier models, the AI market will face a major restructuring. For now, developers are just beginning to explore the new ecosystem, and the outcome of this war will be determined in the coming months.