Original sourceThe DecoderAdditional: Axios
Summary
Benchmark results show Muse Wonder Full scored 56 on the Artificial Analysis Intelligence Index v4.3, with Muse Wonder at 52, both above the previous generation Muse Spark 1.3's 48; on the Coding Agents Index, Wonder Full leads with 68 and Wonder scores 63. Both support a 1 million token context an…
Key points
- Concrete scores and costs let developers make quantitative comparisons with existing models and judge whether Meta's models fit their agent and coding workflows.
- Meta's frontier models have indeed entered the leading group on third-party benchmarks, adding a substantive option for agentic and coding tasks.
- Teams can reconfigure model mixes based on benchmarks and per-task costs, but costs are roughly triple the previous generation—actual usage and budgets need evaluation.
- Compared with the previous generation Muse Spark 1.3, the Muse Wonder series shows a clear improvement on the Intelligence Index.
Editorial note
This page is Code & Chain's editorial summary of public sources. It may be prepared with AI assistance and published through an automated workflow. Refer to the original sources; this content is not investment, legal, or tax advice.