BlogLLM Benchmark Scores as Published, Side by Side

Muse Glimmer-30B: Benchmark Scores as Published

Every result the catalog holds for Muse Glimmer-30B: 8 tests in 6 areas, from 1 page. Each was printed by the company named beside it. Each number is as printed; nothing is ranked or rescored.

Developer MetaReleased 2026-08-09Context window 131KInput Text, imageOutput TextWeights Open (Apache 2.0)Data version 2026103001, as of 2026-10-30

Agentic Coding

TestValuePrinted bySetup as printed
SWE-Bench Pro51.2%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15Harness: Claude Code
Terminal-Bench 2.151.7%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15Harness: Terminus

Science and Math

TestValuePrinted bySetup as printed
GPQA Diamond83.5%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15-

Computer Use

TestValuePrinted bySetup as printed
OSWorld-Verified65.9%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15-

Multimodal Understanding

TestValuePrinted bySetup as printed
CharXiv Reasoning (no tools)78.8%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15-
OmniDocBench 1.575.8%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15-

General Knowledge and Reasoning

TestValuePrinted bySetup as printed
Humanity's Last Exam (no tools)22%Alibaba, Qwen3.8-27B model cardchecked 2026-10-15-

Instruction Following

TestValuePrinted bySetup as printed
IFBench77Alibaba, Qwen3.8-27B model cardchecked 2026-10-15-

More from Meta

Every number here is read from the same catalog file as the benchmark table, data version 2026103001. Where one page printed the model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.