Muse Glimmer 30B

Muse Glimmer 30B is a reasoning model from Meta in the Muse Glimmer family. 25 benchmarks count toward its score, in 7 categories.

availableShows if the model has enough results for an index.
IndexOverall score out of 100.53.5 ±3.7
CoverageShare of the index weight with results.95%
SpeedOutput tokens per second.100/s
Input / 1MUS dollars per 1M input tokens.$0.3 batch $0.175
Output / 1MUS dollars per 1M output tokens.$1.2 batch $0.75 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.131K
EloLMArena rating and rank.N/A

The index is a score out of 100. The ± range shows how much it can change. Batch work costs less.

CapabilitiesScore per category, out of 100.

Out of 100
AgenticMulti-step tasks with tools.
54.1
CodingCode writing and repair.
54.9
ReasoningLogic problems and puzzles.
54.9
MultimodalTasks with images and text.
54.6
KnowledgeFacts and expert knowledge.
48.5
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
55.6
MathMath problems.
54.0

Results

25 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result as a score out of 100.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
AIME 2026Math94.7%54.0Qwen
Artificial Analysis GPQA DiamondKnowledge83.5%54.1Artificial Analysis
Artificial Analysis Long Context ReasoningReasoning83.3%65.9Artificial Analysis
CharXiv ReasoningMultimodal78.8%55.8CharXiv authors
Instruction Following BenchmarkInstruction77.0%55.6Benchmark authors
Software Engineering Benchmark VerifiedCoding76.0%58.5Carlos E. Jimenez et al.
OmniDocBench 1.5Multimodal75.8%OpenAI
MCP AtlasAgentic75.5%65.0OpenAI
ScreenSpot ProMultimodal75.4%57.1Kaixin Li et al.
DeepSearchQAAgentic74.6%57.3Meta AI
Artificial Analysis MMMU-ProMultimodal74.3%57.9Artificial Analysis
Massive Multi-discipline Multimodal Understanding ProMultimodal74.0%47.8MMMU-Pro authors
OSWorld-VerifiedAgentic65.9%55.1Tianbao Xie et al.
Terminal-Bench 2.1 (provider run)Agentic51.7%54.6DeepSeek-AI
SWE-bench ProCoding51.2%53.5Xiang Deng et al.
Artificial Analysis Coding IndexCoding49.0%53.5Artificial Analysis
Artificial Analysis SciCodeCoding44.9%55.2Artificial Analysis
Scientific Code BenchmarkCoding43.6%53.7Benchmark authors
Artificial Analysis EnterpriseOps-GymAgentic34.7%54.9Artificial Analysis
Artificial Analysis Omniscience AccuracyKnowledge27.0%47.3Artificial Analysis
Artificial Analysis Humanity's Last ExamKnowledge22.0%48.4Artificial Analysis
Medical Long Context Reasoning (MLCR-AA)Reasoning20.0%55.1Wisedocs and Artificial Analysis
Artificial Analysis Intelligence IndexKnowledge17.5%44.3Artificial Analysis
GDPval-AA normalizedAgentic13.7%46.1Artificial Analysis
Artificial Analysis Agentic IndexAgentic10.5%45.7Artificial Analysis
Critical Physics TasksReasoning2.6%43.8Artificial Analysis

25 benchmarks count, from 25 of 26 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authorsOpenRouter, collected directlyNo licence stated

Same level, lower price

Qwen3.8-Flash-Next66.4 · Freedots3-note Preview65.5 · FreeApodex 1.1 Mini59.1 · Free

More from Meta

Muse Spark 1.375.0Muse Spark 1.269.2Muse Spark 1.167.0Muse Spark61.6Llama 4 Maverick33.6