Project AURIONAI Use-Cases Research & Innovation
multimodal LLM

Muse Glimmer 30B

Meta's first Apache-licensed open release, a 30B dense agent model; no Llama 5 found. Adapters 13, finetunes 50, merges 1, quants 177.

MetaMetaUSA
Meta
Levelmodel
Classmultimodal LLM
Parameters29.6B
Context131,072
Released10 Aug 2026
Opennessopen weights permissive
Runs onconsumer GPU
Statusactive

Benchmark Results

vendor-reportedmaintainer-run
MCP Atlasagents and tool use
vendor-reported
75.5
050100

Score in %. Higher is better.

huggingface.co
SEA-HELMSoutheast Asian and Indonesian
maintainer-run
75.61 95% CI [-1.13, +1.10]
04080

Score in score. Higher is better.

rank 5, SEA-HELM Indonesian, open-weight models, average of 2,000 bootstraps18 Sep 2026 leaderboard.sea-lion.ai
maintainer-run
72.13 95% CI [-0.59, +0.56]
04080

Score in score. Higher is better.

SEA-HELM SEA average, open-weight models18 Sep 2026 leaderboard.sea-lion.ai
SWE-bench Verifiedcoding
vendor-reported
76
050100

Score in %. Higher is better.

huggingface.co
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

R&D Topics

Agents and tool useproduced byOpen-weight frontier releases as research substrateused by

Identity

Developer
Meta
Family
not in the notes
Released
10 Aug 2026 sourcesnotes
Weights published
–
Record checked
10 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
dense transformer Dense, 52 layers, GQA; DFlash speculative-decoding drafter
Total parameters
29.6B sources
Active parameters
–
Checkpoint parameters
–
Note
Includes 1.8B perception encoder
Context length
131,072 tokens
Input
text, image
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
consumer GPU
Footprint
24 GB GPU at about 17 GB K-quant; under 20 GB at 4-bit; 64 GB full precision
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
measured on a benchmark
  • SEA-HELM Indonesian leaderboard (ID aggregate, first score column of the Indonesian Competency and Task Scores table): 75.61 (95% CI -1.13 to +1.10), run by AI Singapore (SEA-HELM leaderboard maintainer) leaderboard.sea-lion.ai
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Lineage

The notes record no lineage for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. Meta open-sources Muse Glimmer agent model under Apache 2.0 runtimewire.com https://runtimewire.com/article/meta-open-sources-muse-glimmer-agent-model-under-apache-2-0
  2. huggingface.co/meta-models https://huggingface.co/meta-models
  3. huggingface.co/meta-models/Muse-Glimmer-30B https://huggingface.co/meta-models/Muse-Glimmer-30B
  4. SEA-HELM Leaderboard leaderboard.sea-lion.ai/detailed/SEA https://leaderboard.sea-lion.ai/detailed/SEA
  5. SEA-HELM Leaderboard leaderboard.sea-lion.ai/detailed/ID https://leaderboard.sea-lion.ai/detailed/ID

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Muse Glimmer 30B", https://aurion.semai.id/models/meta-muse-glimmer-30b/, records of October 2026.