Muse Spark 1.3
Meta · Released 2 Sep 2026 · muse-spark-1.3
$1.25
Input price / 1M tokens
#27 of 61, cheapest first
$4.25
Output price / 1M tokens
#26 of 61, cheapest first
1.05M
Context window
#11 of 71
152
Output tokens / second
#18 of 61
Summary
Muse Spark 1.3 is a proprietary model from Meta, released on 2 Sep 2026.
At $1.25 input and $4.25 output per million tokens, it is #27 of 61 on input price, cheapest first.
Its 1.05M-token context window ranks #11 of 71.
Measured output speed is 152 tokens per second, #18 of 61.
Meta's current hosted flagship (not a Llama model), sold through the paid Meta Model API. Meta says it is tuned for agentic, long-horizon and coding work, with text, image, video and PDF input.
Benchmarks · 0 reported
No benchmark score we could source for this model. We show scores only with a source link.
Sources
- API ID, modalities, context window, tiers: dev.meta.ai/docs/models, checked 2 Oct 2026
- price, cached price, contributor tier, no long-context premium: dev.meta.ai/docs/pricing-rate-limits, checked 2 Oct 2026
- release date (2 Sep 2026): research.meta.ai/blog/introducing-muse-spark-1-3, checked 2 Oct 2026
- model page and benchmark table: dev.meta.ai/models/muse-spark/, checked 2 Oct 2026
- output speed: artificialanalysis.ai/models/muse-spark-1-3, checked 2 Oct 2026
Building on Muse Spark 1.3?
Get the architecture right before the bill arrives.
We size caching, routing and fallbacks for your workload on a free call, and tell you where a cheaper model is good enough.