Muse Spark
Muse Spark
Meta's closed model, and the clearest sign that Llama is no longer the whole strategy.
Updated
The read
Muse Spark is Meta's proprietary line, separate from Llama and not open in any sense. It performs well, it is deeply wired into Meta's own products, and it exists because the open-weights strategy alone was not winning the frontier. For anyone outside that ecosystem the pitch is thinner: it is a good closed model from a company whose reason to trust it was that it published the weights.
Where it shines
- Competitive with the frontier on general reasoning
- Tightly integrated with Meta's own surfaces and tooling
- Backed by the largest training fleet outside the top two labs
Worth knowing first
- Closed weights from the lab whose reputation was built on open ones
- Thin third-party ecosystem compared with Llama
Index score
81.3
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 20 of 53
- Builds
- LLMs
- Category
- Frontier
- Output
- Text, code and images in, text out
- Pricing
- Paid only
- Best for
- Teams already building inside Meta's platforms
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 01Claude Fable 5Claude Fable 5ProvisionalThe best answer money can buy, and it is a lot of money.Frontier modelPaid only89.2
- 03Claude Opus 5Claude Opus 5ProvisionalMost of the flagship, at half the price. The one to reach for by default.Frontier modelPaid only88.5
- 08Gemini 3.6 FlashGemini 3.6 FlashProvisionalFrontier answers at Flash speed, with a free tier you can actually build on.Frontier modelFree tier86.6
