LMentry
Models perform simple language tasks such as identifying words with a given letter or choosing the longer word. Task-specific rules check the answers.
- Made by
- Avia Efrat and colleagues
- Released
- November 2022
- Kind
- Language reasoning
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- Last model added
- March 2022
- Scores refreshed
- November 20th, 2026
davinci-002 leads with 66.1% among models released by March 2022 davinci-002 leads with 66.1% among models released by March 2022
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
The newest model with a score here came out in March 2022, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 davinci-002 66.1%
- 2 InstructGPT 175B 48.4%
- 3 Tk-dpne 11B 41%
- 4 T0pp 11B 38.3%
- 5 Tk-dp 11B 37.2%
- 6 Tk-d 11B 35.8%
- 7 T0p 11B 35.7%
- 8 InstructGPT 6B 34.3%
- 9 T0 11B 31.6%
- 10 GPT-3 175B (davinci) 29.9%
- 11 InstructGPT 1.3B 24.8%
- 12 GPT-3 6.7B 21.8%
- 13 T5-LM 11B 20.6%
- 14 InstructGPT 350M 20.4%
- 15 GPT-3 XL (babbage) 19.7%
- 16 GPT-3 Medium 15.4%
Source: The LMentry paper
- 1 davinci-002 66.1%
- 2 InstructGPT 175B 48.4%
- 3 InstructGPT 6B 34.3%
- 4 GPT-3 175B (davinci) 29.9%
- 5 InstructGPT 1.3B 24.8%
- 6 GPT-3 6.7B 21.8%
- 7 InstructGPT 350M 20.4%
- 8 GPT-3 XL (babbage) 19.7%
- 9 GPT-3 Medium 15.4%
Source: The LMentry paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | davinci-002 | 66.1% | Mar 2022 | Official leaderboard | |
| 2 | InstructGPT 175B | 48.4% | Jan 2022 | Official leaderboard | |
| 3 | Tk-dpne 11B | 41% | Official leaderboard | ||
| 4 | T0pp 11B | 38.3% | Official leaderboard | ||
| 5 | Tk-dp 11B | 37.2% | Official leaderboard | ||
| 6 | Tk-d 11B | 35.8% | Official leaderboard | ||
| 7 | T0p 11B | 35.7% | Official leaderboard | ||
| 8 | InstructGPT 6B | 34.3% | Jan 2022 | Official leaderboard | |
| 9 | T0 11B | 31.6% | Official leaderboard | ||
| 10 | GPT-3 175B (davinci) | 29.9% | May 2020 | Official leaderboard | |
| 11 | InstructGPT 1.3B | 24.8% | Jan 2022 | Official leaderboard | |
| 12 | GPT-3 6.7B | 21.8% | Jun 2020 | Official leaderboard | |
| 13 | T5-LM 11B | 20.6% | Official leaderboard | ||
| 14 | InstructGPT 350M | 20.4% | Jan 2022 | Official leaderboard | |
| 15 | GPT-3 XL (babbage) | 19.7% | Jun 2020 | Official leaderboard | |
| 16 | GPT-3 Medium | 15.4% | Jun 2020 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | davinci-002 | 66.1% | Mar 2022 | Official leaderboard | |
| 2 | InstructGPT 175B | 48.4% | Jan 2022 | Official leaderboard | |
| 3 | InstructGPT 6B | 34.3% | Jan 2022 | Official leaderboard | |
| 4 | GPT-3 175B (davinci) | 29.9% | May 2020 | Official leaderboard | |
| 5 | InstructGPT 1.3B | 24.8% | Jan 2022 | Official leaderboard | |
| 6 | GPT-3 6.7B | 21.8% | Jun 2020 | Official leaderboard | |
| 7 | InstructGPT 350M | 20.4% | Jan 2022 | Official leaderboard | |
| 8 | GPT-3 XL (babbage) | 19.7% | Jun 2020 | Official leaderboard | |
| 9 | GPT-3 Medium | 15.4% | Jun 2020 | Official leaderboard |
Rankings by part
We read 25 parts from the benchmark’s paper, each ranking models on one part of the test, such as sentence containing, sentence not containing and word containing. davinci-002 leads 22 of them, and InstructGPT 175B leads the other 3. Arrows show how far a model moved from its overall rank among the models we track.
sentence containing
9 models we track have a score here.
- 1 davinci-002 96.1%
- 2 InstructGPT 175B 95.5%
- 3 GPT-3 175B (davinci) ↑1 90.4%
- 4 InstructGPT 6B ↓1 87.5%
- 5 InstructGPT 350M ↑2 81.8%
- 6 InstructGPT 1.3B ↓1 79.9%
- 7 GPT-3 XL (babbage) ↑1 75.5%
- 8 GPT-3 6.7B ↓2 74.3%
- 9 GPT-3 Medium 72%
sentence not containing
9 models we track have a score here.
- 1 davinci-002 90.4%
- 2 InstructGPT 175B 81.8%
- 3 InstructGPT 1.3B ↑2 36.6%
- 4 GPT-3 XL (babbage) ↑4 34.7%
- 5 GPT-3 Medium ↑4 32.6%
- 6 InstructGPT 350M ↑1 32.4%
- 7 GPT-3 175B (davinci) ↓3 25.1%
- 8 InstructGPT 6B ↓5 18.1%
- 9 GPT-3 6.7B ↓3 11.6%
word containing
9 models we track have a score here.
- 1 InstructGPT 175B ↑1 100%
- 2 InstructGPT 6B ↑1 98.7%
- 3 InstructGPT 1.3B ↑2 96.2%
- 4 davinci-002 ↓3 85.9%
- 5 InstructGPT 350M ↑2 78.2%
- 6 GPT-3 XL (babbage) ↑2 29.5%
- 7 GPT-3 175B (davinci) ↓3 11.5%
- 8 GPT-3 6.7B ↓2 9.00%
- 9 GPT-3 Medium 7.70%
word not containing
9 models we track have a score here.
- 1 InstructGPT 175B ↑1 78.2%
- 2 davinci-002 ↓1 38.5%
- 3 InstructGPT 1.3B ↑2 16.7%
- 4 InstructGPT 350M ↑3 12.8%
- 5 InstructGPT 6B ↓2 12.8%
- 6 GPT-3 XL (babbage) ↑2 11.5%
- 7 GPT-3 175B (davinci) ↓3 0%
- 8 GPT-3 6.7B ↓2 0%
- 9 GPT-3 Medium 0%
most associated word
9 models we track have a score here.
- 1 davinci-002 97.5%
- 2 InstructGPT 175B 47.2%
- 3 GPT-3 175B (davinci) ↑1 26.7%
- 4 InstructGPT 6B ↓1 25.5%
- 5 GPT-3 6.7B ↑1 20.6%
- 6 InstructGPT 1.3B ↓1 20.1%
- 7 GPT-3 Medium ↑2 14.2%
- 8 InstructGPT 350M ↓1 13%
- 9 GPT-3 XL (babbage) ↓1 6.40%
least associated word
9 models we track have a score here.
- 1 davinci-002 97.8%
- 2 InstructGPT 175B 57.6%
- 3 GPT-3 175B (davinci) ↑1 32.9%
- 4 GPT-3 6.7B ↑2 7.70%
- 5 InstructGPT 6B ↓2 6.30%
- 6 InstructGPT 1.3B ↓1 4.30%
- 7 GPT-3 XL (babbage) ↑1 3.30%
- 8 GPT-3 Medium ↑1 0.70%
- 9 InstructGPT 350M ↓2 0.70%
any words from category
9 models we track have a score here.
- 1 davinci-002 58.2%
- 2 InstructGPT 6B ↑1 57.5%
- 3 GPT-3 175B (davinci) ↑1 54%
- 4 InstructGPT 350M ↑3 52.2%
- 5 InstructGPT 175B ↓3 52%
- 6 GPT-3 Medium ↑3 50.2%
- 7 GPT-3 XL (babbage) ↑1 49.6%
- 8 GPT-3 6.7B ↓2 48.5%
- 9 InstructGPT 1.3B ↓4 47.9%
all words from category
9 models we track have a score here.
- 1 davinci-002 69.1%
- 2 InstructGPT 175B 54.4%
- 3 GPT-3 Medium ↑6 51.9%
- 4 GPT-3 6.7B ↑2 50.2%
- 5 InstructGPT 350M ↑2 49.7%
- 6 GPT-3 175B (davinci) ↓2 49.4%
- 7 GPT-3 XL (babbage) ↑1 49.4%
- 8 InstructGPT 6B ↓5 48.9%
- 9 InstructGPT 1.3B ↓4 46.1%
first alphabetically
9 models we track have a score here.
- 1 davinci-002 73.1%
- 2 InstructGPT 175B 53.8%
- 3 GPT-3 175B (davinci) ↑1 51.2%
- 4 InstructGPT 6B ↓1 49.3%
- 5 InstructGPT 1.3B 48.7%
- 6 GPT-3 6.7B 46.1%
- 7 GPT-3 Medium ↑2 45.1%
- 8 GPT-3 XL (babbage) 38.3%
- 9 InstructGPT 350M ↓2 24.2%
more letters
9 models we track have a score here.
- 1 davinci-002 69.2%
- 2 InstructGPT 6B ↑1 55.3%
- 3 InstructGPT 175B ↓1 53.7%
- 4 GPT-3 Medium ↑5 52.3%
- 5 GPT-3 6.7B ↑1 52.2%
- 6 GPT-3 XL (babbage) ↑2 51.2%
- 7 InstructGPT 1.3B ↓2 50.8%
- 8 GPT-3 175B (davinci) ↓4 47.8%
- 9 InstructGPT 350M ↓2 26.9%
less letters
9 models we track have a score here.
- 1 davinci-002 68.6%
- 2 InstructGPT 175B 49.9%
- 3 GPT-3 6.7B ↑3 47.1%
- 4 InstructGPT 6B ↓1 44.5%
- 5 GPT-3 XL (babbage) ↑3 44%
- 6 GPT-3 175B (davinci) ↓2 39.4%
- 7 InstructGPT 1.3B ↓2 38.2%
- 8 GPT-3 Medium ↑1 38.1%
- 9 InstructGPT 350M ↓2 25.6%
bigger number
9 models we track have a score here.
- 1 davinci-002 97.6%
- 2 InstructGPT 175B 76.3%
- 3 InstructGPT 6B 58.1%
- 4 GPT-3 175B (davinci) 56.9%
- 5 GPT-3 6.7B ↑1 36.6%
- 6 InstructGPT 1.3B ↓1 32.4%
- 7 GPT-3 XL (babbage) ↑1 27.6%
- 8 GPT-3 Medium ↑1 25.3%
- 9 InstructGPT 350M ↓2 17.3%
smaller number
9 models we track have a score here.
- 1 davinci-002 98.7%
- 2 InstructGPT 175B 79.8%
- 3 InstructGPT 6B 64%
- 4 GPT-3 175B (davinci) 38.7%
- 5 InstructGPT 1.3B 36.6%
- 6 GPT-3 6.7B 33.5%
- 7 InstructGPT 350M 27.1%
- 8 GPT-3 XL (babbage) 20.3%
- 9 GPT-3 Medium 11.2%
rhyming word
9 models we track have a score here.
- 1 davinci-002 66.7%
- 2 InstructGPT 175B 41.9%
- 3 GPT-3 175B (davinci) ↑1 22.4%
- 4 InstructGPT 1.3B ↑1 21.8%
- 5 GPT-3 6.7B ↑1 12.7%
- 6 InstructGPT 6B ↓3 11.9%
- 7 GPT-3 Medium ↑2 10.3%
- 8 GPT-3 XL (babbage) 9.40%
- 9 InstructGPT 350M ↓2 7.80%
homophones
9 models we track have a score here.
- 1 davinci-002 72.2%
- 2 InstructGPT 175B 56.4%
- 3 GPT-3 175B (davinci) ↑1 48.4%
- 4 InstructGPT 6B ↓1 25.2%
- 5 InstructGPT 1.3B 21.6%
- 6 GPT-3 6.7B 18.7%
- 7 GPT-3 Medium ↑2 11.1%
- 8 GPT-3 XL (babbage) 6.10%
- 9 InstructGPT 350M ↓2 6.10%
word after
9 models we track have a score here.
- 1 davinci-002 90.6%
- 2 InstructGPT 175B 55%
- 3 GPT-3 175B (davinci) ↑1 47%
- 4 InstructGPT 6B ↓1 27.5%
- 5 GPT-3 6.7B ↑1 23.9%
- 6 InstructGPT 1.3B ↓1 9.70%
- 7 GPT-3 XL (babbage) ↑1 7.10%
- 8 InstructGPT 350M ↓1 1.60%
- 9 GPT-3 Medium 0.30%
word before
9 models we track have a score here.
- 1 davinci-002 48.8%
- 2 InstructGPT 175B 21.3%
- 3 GPT-3 175B (davinci) ↑1 12.3%
- 4 GPT-3 6.7B ↑2 10.7%
- 5 InstructGPT 6B ↓2 8.40%
- 6 GPT-3 XL (babbage) ↑2 5.50%
- 7 InstructGPT 1.3B ↓2 3.30%
- 8 InstructGPT 350M ↓1 2.40%
- 9 GPT-3 Medium 0.40%
starts with word
9 models we track have a score here.
- 1 davinci-002 80%
- 2 GPT-3 175B (davinci) ↑2 63.2%
- 3 GPT-3 XL (babbage) ↑5 30.4%
- 4 GPT-3 Medium ↑5 23.6%
- 5 InstructGPT 175B ↓3 23.2%
- 6 GPT-3 6.7B 15.3%
- 7 InstructGPT 6B ↓4 8.10%
- 8 InstructGPT 1.3B ↓3 5.20%
- 9 InstructGPT 350M ↓2 4.00%
ends with word
9 models we track have a score here.
- 1 davinci-002 42.5%
- 2 InstructGPT 175B 20.3%
- 3 GPT-3 175B (davinci) ↑1 13.2%
- 4 InstructGPT 6B ↓1 12.8%
- 5 GPT-3 Medium ↑4 9.50%
- 6 InstructGPT 1.3B ↓1 8.50%
- 7 GPT-3 6.7B ↓1 8.30%
- 8 GPT-3 XL (babbage) 7.90%
- 9 InstructGPT 350M ↓2 7.40%
starts with letter
9 models we track have a score here.
- 1 InstructGPT 175B ↑1 100%
- 2 davinci-002 ↓1 100%
- 3 InstructGPT 1.3B ↑2 89.7%
- 4 InstructGPT 350M ↑3 88.5%
- 5 InstructGPT 6B ↓2 80.8%
- 6 GPT-3 XL (babbage) ↑2 20.5%
- 7 GPT-3 6.7B ↓1 16.7%
- 8 GPT-3 175B (davinci) ↓4 15.4%
- 9 GPT-3 Medium 6.40%
ends with letter
9 models we track have a score here.
- 1 davinci-002 84.1%
- 2 InstructGPT 175B 62.3%
- 3 GPT-3 XL (babbage) ↑5 27.5%
- 4 InstructGPT 6B ↓1 21.7%
- 5 InstructGPT 1.3B 20.3%
- 6 InstructGPT 350M ↑1 7.20%
- 7 GPT-3 Medium ↑2 5.80%
- 8 GPT-3 6.7B ↓2 4.40%
- 9 GPT-3 175B (davinci) ↓5 1.40%
first word
9 models we track have a score here.
- 1 davinci-002 98.5%
- 2 InstructGPT 175B 94.6%
- 3 GPT-3 175B (davinci) ↑1 63.7%
- 4 InstructGPT 6B ↓1 59%
- 5 GPT-3 6.7B ↑1 41.4%
- 6 GPT-3 XL (babbage) ↑2 27.3%
- 7 InstructGPT 350M 16.1%
- 8 InstructGPT 1.3B ↓3 12.5%
- 9 GPT-3 Medium 8.40%
last word
9 models we track have a score here.
- 1 davinci-002 93.8%
- 2 InstructGPT 175B 60.8%
- 3 GPT-3 175B (davinci) ↑1 43.3%
- 4 InstructGPT 6B ↓1 32.4%
- 5 GPT-3 6.7B ↑1 22.1%
- 6 GPT-3 XL (babbage) ↑2 12.5%
- 7 InstructGPT 1.3B ↓2 10%
- 8 InstructGPT 350M ↓1 5.60%
- 9 GPT-3 Medium 5.50%
first letter
9 models we track have a score here.
- 1 davinci-002 99.9%
- 2 InstructGPT 175B 99.8%
- 3 InstructGPT 6B 88.1%
- 4 InstructGPT 1.3B ↑1 82.6%
- 5 GPT-3 175B (davinci) ↓1 71.3%
- 6 GPT-3 6.7B 69.6%
- 7 GPT-3 XL (babbage) ↑1 44.7%
- 8 InstructGPT 350M ↓1 38.2%
- 9 GPT-3 Medium 23.9%
last letter
9 models we track have a score here.
- 1 davinci-002 98.5%
- 2 InstructGPT 175B 62.9%
- 3 InstructGPT 6B 41.1%
- 4 GPT-3 175B (davinci) 23.8%
- 5 GPT-3 6.7B ↑1 18.9%
- 6 InstructGPT 1.3B ↓1 15.4%
- 7 InstructGPT 350M 3.20%
- 8 GPT-3 XL (babbage) 2.10%
- 9 GPT-3 Medium 0.90%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The LMentry paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official LMentry 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
- 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 4 years ago, in March 2022, for davinci-002. 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 benchmark’s paper (16 entries).
- Pass Models scored 9 models we track, plus 7 other entries on its leaderboard.
- Warning Newest score The newest came 4 years ago, in March 2022, for davinci-002.
- Note Frontier models None of its models are on the capabilities index.
- Pass Headroom The best is 66.1%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on aclanthology.org: 9 rows for models we track and 7 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.