Benchmark Almanac

Math23K

GLM-4 88.5% leads since June 2024
Only in system cards

Models solve Chinese elementary-school math word problems from a set of about 23,000. Scored by final-answer accuracy.

Introduced with the Deep Neural Solver paper at EMNLP 2017.

Released
September 2017
Kind
Word problems
Level
School
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Strongest model tested
Qwen-14B #213 on the capability index
Last in a system card
June 2024
Last model added
June 2024
3
Models scored in lab cards
3
System cards report it, from 2 labs

GLM-4 leads with 88.5% among models released by January 2024

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

The newest model with a score here came out in January 2024, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 88.5 GLM-4
  2. 67.0 Qwen-14B
  3. 51.2 Qwen-7B
  1. 1 GLM-4 88.5%
  2. 2 Qwen-14B 67%
  3. 3 Qwen-7B 51.2%

Source: Labs’ model cards

#ModelScoreSettingDateSource
1 GLM-4 88.5% Jun 2024 Lab model card
2 Qwen-14B 67% Sep 2023 Lab model card
3 Qwen-7B 51.2% Sep 2023 Lab model card

2 models have held number one in labs’ own numbers since 2023

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

Qwen Z.ai
Lab claims Passed 85% in Jun 2024, by GLM-4
GLM-4 88.5%

Source: Labs’ model cards

Z.ai’s own number for GLM-4 passed 85% in June 2024

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

Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) GLM-4 Qwen-14B 0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) GLM-4 Qwen-14B

Source: Labs’ model cards

Labs report it only in system cards

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.

3 system cards report it, from 2 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Jun 18, 2024 GLM-4 technical report 88.5 GLM-4 All Tools (Web, 0116), Python Interpreter, Math23K (Table 9) Capabilities
Sep 28, 2023 Qwen-7B technical report 51.2 Math23K, greedy decoding (Table 12) Capabilities
Sep 28, 2023 Qwen-14B technical report 67 Math23K, greedy decoding (Table 12) Capabilities

More in Math

All 113 in Math
  • AIME 2026 In use Models solve the 30 integer-answer problems from the 2026 AIME I and II math competitions. Scored by exact answer match. GPT-5.5 100%
  • IMO-AnswerBench In use 400 olympiad problems with short verifiable answers, robustified against memorization. Scored by answer match. GLM-5.2 91%
  • HMMT 2026 In use Models solve the 33 final-answer problems from the February 2026 Harvard-MIT Mathematics Tournament. Scored by exact answer match. GPT-5.5 98.5%
  • PolyMath In use Models solve 9,000 parallel math problems in 18 languages across four difficulty levels, from school math to frontier math. Final answers are checked. Qwen3.7-Max 86.5%
  • MathArena (Apex, IMO, USAMO) In use Evaluates models on newly released math competitions after release. Final-answer or human/LLM-judged proof scores. Claude Opus 4.8 81.3%
  • MathArena Apex Shortlist In use Models solve 47 hard final-answer problems that MathArena selected from 2025 competitions, where the best models scored about 50% at selection. Scored by accuracy over repeated attempts. GPT-5.5 98.4%