PyyanThe AI Index

Pyyan / Compare / Muse Spark Open vs MiMo-V2.6-Flash

Muse Spark Open vs MiMo-V2.6-Flash

2 of 5

Open-Weight Models · verified 20 Oct 2026

Muse Spark OpenMetacurrent
XMiMo-V2.6-FlashXiaomicurrent
3 slots left
SpecificationMuse Spark OpenMiMo-V2.6-Flash
SummaryThe largest American open weights, under a licence with a revenue trigger.The small one, and still multimodal.
Context1M1M
Input ($/Mtok)Weights onlyWeights only
Output ($/Mtok)Weights onlyWeights only
LicenceMuse Model LicenseMIT
Max outputNot publishedNot published
Parameters816B total, 42B active309B total, 15B active
Released15 Oct 202621 Sep 2026
Model IDmeta-models/Muse-Spark-Open-816B-A42BXiaomiMiMo/MiMo-V2.6-Flash
CategoryOpen-Weight ModelsOpen-Weight Models
OfficialMeta ↗Xiaomi ↗

Highlighted rows are where these differ.

Muse Spark Open

  • Text, image, video and audio in, text out, with a 1M token context
  • Muse Model License, not open source: attribution required, and a separate agreement above $20m of model-service revenue
  • About 1.5TB in BF16, or two Mac Studios at 4-bit

Best for multimodal work you host yourself.

Full spec sheet →

MiMo-V2.6-Flash

  • 309B total parameters, 15B active per token
  • Same 1M context and five layer speculative decoder as Pro
  • MIT licence, weights on Hugging Face

Best for multimodal work on your own cluster.

Full spec sheet →