Benchmark Almanac

MusicTheoryBench

gemini-3.1-flash 72.2% tops its leaderboard Claude Sonnet 4.6 71.1% leads the models we track
Not in model cards

Models answer college music theory questions about musical knowledge and reasoning. Some questions include a score written in ABC notation.

MuTheoryEval ran 15 configurations on all 367 test questions. It scored generated answers. The ChatMusician authors scored answer choices by perplexity.

Released
December 2023
Kind
Exam questions
Level
University
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Subject
Arts
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Exams and contests
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-5.4 #28 on the capability index
Last model added
March 2026
Scores refreshed
November 20th, 2026
13
Entries on MuTheoryEval’s leaderboard, 8 models we track

gemini-3.1-flash tops the leaderboard with 72.2% Claude Sonnet 4.6 leads with 71.1% among models released by March 2026

Best score per entry from MuTheoryEval’s leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Claude Sonnet 4.6, at 71.1%. This view shows only models we track, each at its best.

The newest model with a score here came out in March 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 gemini-3.1-flash lite, temperature=0, max_output_tokens=4096, thinking={"level":"default"} 72.2%
  2. 2 gemini-3.1-flash (minimal) lite, temperature=0, max_output_tokens=4096, thinking={"level":"minimal"}, answer_format=MCQ 72.2%
  3. 3 Claude Sonnet 4.6 lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ 71.1%
  4. 4 Claude Opus 4.6 lite, temperature=0, max_output_tokens=16, thinking=false 67.6%
  5. 5 Gemini 3.1 Flash-Lite lite 66.8%
  6. 6 GLM-5 lite, temperature=0, max_output_tokens=1024, thinking=false 64.3%
  7. 7 Qwen3-Max lite, max_output_tokens=1024, thinking=native 63.2%
  8. 8 GPT-5.4 lite 63.2%
  9. 9 deepseek-chat lite, temperature=0, max_output_tokens=16, thinking=false 61.6%
  10. 10 GPT-5.4 Mini lite, max_output_tokens=1024, thinking=native, answer_format=MCQ 59.4%
  11. 11 Claude Haiku 4.5 lite, temperature=0, max_output_tokens=16, thinking=false, answer_format=MCQ 58.9%
  12. 12 glm-z1-flash lite, max_output_tokens=4096, thinking=native 48.2%
  13. 13 deepseek-reasoner lite, max_output_tokens=16, thinking=native 41.7%

Source: MuTheoryEval’s leaderboard

#EntryScoreSettingReleasedRun onSource
1 gemini-3.1-flash 72.2% lite, temperature=0, max_output_tokens=4096, thinking={"level":"default"} Apr 2026 Official leaderboard
2 gemini-3.1-flash (minimal) 72.2% lite, temperature=0, max_output_tokens=4096, thinking={"level":"minimal"}, answer_format=MCQ Apr 2026 Official leaderboard
3 Claude Sonnet 4.6 71.1% lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ Feb 2026 Apr 2026 Official leaderboard
4 Claude Opus 4.6 67.6% lite, temperature=0, max_output_tokens=16, thinking=false Feb 2026 Apr 2026 Official leaderboard
5 Gemini 3.1 Flash-Lite 66.8% lite Mar 2026 Apr 2026 Official leaderboard
6 GLM-5 64.3% lite, temperature=0, max_output_tokens=1024, thinking=false Feb 2026 Apr 2026 Official leaderboard
7 Qwen3-Max 63.2% lite, max_output_tokens=1024, thinking=native Sep 2025 Apr 2026 Official leaderboard
8 GPT-5.4 63.2% lite Mar 2026 Apr 2026 Official leaderboard
9 deepseek-chat 61.6% lite, temperature=0, max_output_tokens=16, thinking=false Apr 2026 Official leaderboard
10 GPT-5.4 Mini 59.4% lite, max_output_tokens=1024, thinking=native, answer_format=MCQ Mar 2026 Apr 2026 Official leaderboard
11 Claude Haiku 4.5 58.9% lite, temperature=0, max_output_tokens=16, thinking=false, answer_format=MCQ Oct 2025 Apr 2026 Official leaderboard
12 glm-z1-flash 48.2% lite, max_output_tokens=4096, thinking=native Apr 2026 Official leaderboard
13 deepseek-reasoner 41.7% lite, max_output_tokens=16, thinking=native Apr 2026 Official leaderboard
#ModelScoreSettingReleasedRun onSource
1 Claude Sonnet 4.6 71.1% lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ Feb 2026 Apr 2026 Official leaderboard
2 Claude Opus 4.6 67.6% lite, temperature=0, max_output_tokens=16, thinking=false Feb 2026 Apr 2026 Official leaderboard
3 Gemini 3.1 Flash-Lite 66.8% lite Mar 2026 Apr 2026 Official leaderboard
4 GLM-5 64.3% lite, temperature=0, max_output_tokens=1024, thinking=false Feb 2026 Apr 2026 Official leaderboard
5 Qwen3-Max 63.2% lite, max_output_tokens=1024, thinking=native Sep 2025 Apr 2026 Official leaderboard
6 GPT-5.4 63.2% lite Mar 2026 Apr 2026 Official leaderboard
7 GPT-5.4 Mini 59.4% lite, max_output_tokens=1024, thinking=native, answer_format=MCQ Mar 2026 Apr 2026 Official leaderboard
8 Claude Haiku 4.5 58.9% lite, temperature=0, max_output_tokens=16, thinking=false, answer_format=MCQ Oct 2025 Apr 2026 Official leaderboard

Rankings by part

We read 2 parts from MuTheoryEval’s leaderboard: knowledge and reasoning. Each ranks models on one part of the test. GPT-5.4 Mini leads knowledge and Claude Haiku 4.5 leads reasoning. Arrows show how far a model moved from its overall rank among the models we track.

knowledge

3 models we track have a score here.

  1. 1 GPT-5.4 Mini ↑6 69.9%
  2. 2 GLM-5 ↑2 68.8%
  3. 3 Claude Haiku 4.5 ↑5 66.9%

The best score, 71.1%, costs $0.06 for the full run

Runs on the cost frontier score higher than any run that's cheaper. Gemini 3.1 Flash-Lite at lite comes within 4.4 points of the best for $0.006 for the full run, 10 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 Cost of the full run, log scale Claude Sonnet 4.6, lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ Gemini 3.1 Flash-Lite, lite 0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 Cost of the full run, log scale Claude Sonnet 4.6, lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ

Source: MuTheoryEval’s leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
Claude Sonnet 4.6 lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ 71.1% $0.06
Claude Opus 4.6 lite, temperature=0, max_output_tokens=16, thinking=false 67.6% $0.30 4
Gemini 3.1 Flash-Lite lite 66.8% $0.006
GLM-5 lite, temperature=0, max_output_tokens=1024, thinking=false 64.3% $0.06 2
Qwen3-Max lite, max_output_tokens=1024, thinking=native 63.2% $0.06 1
GPT-5.4 lite 63.2% $0.15
Claude Sonnet 4.6 lite, temperature=0, max_output_tokens=16, thinking=false 62.4% $0.18 4
GPT-5.4 Mini lite, max_output_tokens=1024, thinking=native, answer_format=MCQ 59.4% $0.05 5
Claude Haiku 4.5 lite, temperature=0, max_output_tokens=16, thinking=false, answer_format=MCQ 58.9% $0.06 4
GLM-5 lite, max_output_tokens=4096, thinking=enabled, answer_format=MCQ 57.2% $0.06 1,108

The best score, 71.1%, costs $0.06 for the full run

Runs on the cost frontier score higher than any run that's cheaper. Gemini 3.1 Flash-Lite at lite comes within 4.4 points of the best for $0.006 for the full run, 10 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 Cost of the full run, log scale Claude Sonnet 4.6, lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ Gemini 3.1 Flash-Lite, lite 0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 Cost of the full run, log scale Claude Sonnet 4.6, lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ

Source: MuTheoryEval’s leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
Claude Sonnet 4.6 lite, temperature=0, max_output_tokens=8256, thinking={"budget_tokens":8000}, answer_format=MCQ 71.1% $0.06
Claude Opus 4.6 lite, temperature=0, max_output_tokens=16, thinking=false 67.6% $0.30 4
Gemini 3.1 Flash-Lite lite 66.8% $0.006
GLM-5 lite, temperature=0, max_output_tokens=1024, thinking=false 64.3% $0.06 2
Qwen3-Max lite, max_output_tokens=1024, thinking=native 63.2% $0.06 1
GPT-5.4 lite 63.2% $0.15
Claude Sonnet 4.6 lite, temperature=0, max_output_tokens=16, thinking=false 62.4% $0.18 4
GPT-5.4 Mini lite, max_output_tokens=1024, thinking=native, answer_format=MCQ 59.4% $0.05 5
Claude Haiku 4.5 lite, temperature=0, max_output_tokens=16, thinking=false, answer_format=MCQ 58.9% $0.06 4
GLM-5 lite, max_output_tokens=4096, thinking=enabled, answer_format=MCQ 57.2% $0.06 1,108

3 models have held number one on MuTheoryEval’s leaderboard since its release in December 2023

Lab claims count from the day the card came out, independent runs from the model’s release.

Qwen Anthropic
MuTheoryEval leaderboard Not tested since Mar 2026

Source: MuTheoryEval’s leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On MuTheoryEval’s leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Sonnet 4.6 Qwen3-Max 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Sonnet 4.6 Qwen3-Max

Source: MuTheoryEval’s leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Knowledge

All 332 in Knowledge
  • MMLU-Pro In use Harder 10-option version of MMLU with reasoning-focused questions across 14 subjects. Intern-S2-397B 89.8%
  • MMLU In use 57-subject multiple-choice exam questions from elementary to professional level. Scored by accuracy. Seed1.8 92.3%
  • SuperGPQA In use 26,529 graduate-level multiple-choice questions across 285 disciplines. Accuracy. Qwen 3.6 Max (Preview) 73.9%
  • WideSearch In use 200 bilingual broad info-seeking tasks: agent collects many atomic facts into a table. Scored by success rate and row/item F1. Tencent Hy4 preview 83.9%
  • BBEH micro average In use Models solve harder replacements for each of the 23 BIG-Bench Hard tasks, 4,520 examples in total. The score is the share of all examples answered correctly. Gemma 4 31B IT 74.4%
  • HLE-Diamond In use Models answer 1,000 cleaned expert questions from Humanity's Last Exam, 500 on reasoning and 500 on knowledge, closed book. Graded on exact answers. Claude Fable 5.2 77.9%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 8 months ago, in March 2026, for GPT-5.4 Mini. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (13 entries).
  • Pass Models scored 8 models we track, plus 5 other entries on its leaderboard.
  • Warning Newest score The newest came 8 months ago, in March 2026, for GPT-5.4 Mini.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.4 (#28).
  • Pass Headroom The best is 71.1%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on MuTheoryEval: 10 rows for models we track and 5 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check