Quick summary
Taken from the multiple choice board, the board with the most models in this set. The other boards have their own figures below.
Results
How many questions each model answered correctly, from the exact same set. The bars show the count of correct answers, not a general ranking of model ability.
The model reads the question with its answer options, then replies with a single letter. Scoring is automatic: correct when the letter matches the key.
A model whose name does not end in a date is called through an id that does not pin the version. The provider may swap the model behind that id at any time without notice, so the figures apply to whichever version was active when the run happened.
How these figures are computed
mcq-py1KeteranganHow multiple choice is asked. Question and options are built from the item, option order is shuffled with a fixed seed so every model sees the same order, and the model is asked for a single letter.scoring mcq-py1KeteranganHow multiple choice is scored. We read one option letter from the answer and compare it with the key. An answer with no option letter is recorded as a format failure, kept apart from a wrong answer, so a broken harness does not read as a weak model.question version c0c83850858eKeteranganA 12-character fingerprint of every question the model saw: item number, question text, answer options, and option order. Changing a single word changes the fingerprint. Scores under different fingerprints answer different questions, so this page only joins figures that share one.answer key version 18e8ec378904KeteranganA 12-character fingerprint of the answer key: the reference, the other answers we accept, and speaker verdicts for items no machine may score. The questions can stay identical while the key changes, and when that happens scores move without a single word of the question changing. That is why it is kept apart from the question version.The model writes its own answer, with no options offered. Some are scored automatically against a reference, some are read one by one by a native speaker using a weighted rubric.
A model whose name does not end in a date is called through an id that does not pin the version. The provider may swap the model behind that id at any time without notice, so the figures apply to whichever version was active when the run happened.
How these figures are computed
open-py1KeteranganThe prompt reads “Rewrite the sentence in standard Indonesian”, taken as it stands from inside the item. It is loose: it never says what must be left alone, so a model is free to read it as licence to swap words for synonyms.scoring open-py1KeteranganHow open answers are scored. The answer is normalised first, then compared with the reference and the list of other answers we accept. Items that require human judgement are deliberately left unscored.question version 9f2d79048c32KeteranganA 12-character fingerprint of every question the model saw: item number, question text, answer options, and option order. Changing a single word changes the fingerprint. Scores under different fingerprints answer different questions, so this page only joins figures that share one.answer key version 4610a4b37355KeteranganA 12-character fingerprint of the answer key: the reference, the other answers we accept, and speaker verdicts for items no machine may score. The questions can stay identical while the key changes, and when that happens scores move without a single word of the question changing. That is why it is kept apart from the question version.Per-model detail
The same figures as the chart above, plus the cost per correct answer, the wait time, and the date of the last run. Any column that cannot be guessed from its name carries its own explanation.
| Model | CorrectKeteranganHow many answers matched the key, out of the items scored. | Held backKeteranganThe score on items whose key we have never published. The column to its left covers items whose key is public, so a score there cannot be told apart from a model that read it. This column is what tells them apart. The held-back slice on this set is too small to read as a measurement: a gap only means something above roughly 44.8 points, far larger than any realistic gap. The percentage is therefore deliberately not shown, only the count. Publishing this score also wears it down over time, because every figure released is information about the items we hold back. That is why the held-back slice has a limited lifespan, and will be rotated. | $/correctKeteranganThe cost of one full run divided by ALL its correct answers, including those in the Held back column. The Correct column shows the public slice only, so the denominator here is the sum of both columns. NOT the cost per call: a cheap model that gets a lot wrong can cost more per correct answer than a model that costs more per call. | LatencyKeteranganMedian wait for a single call. Median rather than mean, so one stalled call does not move the number. | Last runKeteranganThe last date we ran this model on the same questions and answer key. |
|---|---|---|---|---|---|
| Gemini 3.5 FlashGoogle | 26/2796.3% | 11/11 | $0.002148$0.0795 | 2,217 msmedian | |
| Gemma 4 31BGoogle | 26/2796.3% | 10/11 | $0.000015$0.0006 | 803 msmedian | |
| DeepSeek V4 Flash 0731DeepSeek | 25/2792.6% | 11/11 | $0.000054$0.0020 | 4,324 msmedian | |
| GPT-5.6 LunaOpenAI | 23/2785.2% | 11/11 | $0.000023$0.0008 | 988 msmedian | |
| Claude Haiku 4.5Anthropic | 20/2774.1% | 8/11 | $0.000256$0.0072 | 2,228 msmedian | |
| Ling-3.0-flashAnt Group | 18/2766.7% | 8/11 | $0.000058$0.0015 | 1,547 msmedian | |
| Solar Pro 4Upstage | 16/2759.3% | 6/11 | $0.000010$0.0002 | 6,680 msmedian | |
| Gemma-SEA-LION v4.5 E2B-ITlocal | 16/2759.3% | 6/11 | - | 837 msmedian | |
| Qwen3 30B A3B Instruct 2507Alibaba | 14/2751.9% | 5/11 | $0.000023$0.0004 | 612 msmedian | |
| Llama3 8B CPT Sahabat-AI v1 Instructlocal | 13/2748.1% | 5/11 | - | 311 msmedian | |
| Hunyuan A13B InstructTencent | 12/2744.4% | 9/11 | $0.000039$0.0008 | 1,396 msmedian | |
| Mistral Small 3Mistral | 11/2740.7% | 7/11 | $0.000014$0.0003 | 547 msmedian | |
| Llama 3.1 8B Instructlocal | 9/2733.3% | 6/11 | - | 432 msmedian |
Models marked local run on our own hardware, so there is no bill to record and the wait time measures our machine, not a service anyone else can buy. The score is still comparable: the items, the key, and the temperature are identical to every other row here.
| Model | CorrectKeteranganHow many answers matched the key, out of the items scored. | Held backKeteranganThe score on items whose key we have never published. The column to its left covers items whose key is public, so a score there cannot be told apart from a model that read it. This column is what tells them apart. The held-back slice on this set is too small to read as a measurement: a gap only means something above roughly 75.3 points, far larger than any realistic gap. The percentage is therefore deliberately not shown, only the count. Publishing this score also wears it down over time, because every figure released is information about the items we hold back. That is why the held-back slice has a limited lifespan, and will be rotated. | $/correctKeteranganThe cost of one full run divided by ALL its correct answers, including those in the Held back column. The Correct column shows the public slice only, so the denominator here is the sum of both columns. NOT the cost per call: a cheap model that gets a lot wrong can cost more per correct answer than a model that costs more per call. | LatencyKeteranganMedian wait for a single call. Median rather than mean, so one stalled call does not move the number. | Last runKeteranganThe last date we ran this model on the same questions and answer key. |
|---|---|---|---|---|---|
| Gemma 4 31BGoogle | 6/6100.0% | 2/2 | $0.000019$0.0001 | 1,318 msmedian | |
| Gemini 3.5 FlashGoogle | 6/6100.0% | 2/2 | $0.003273$0.0262 | 2,753 msmedian | |
| Gemma-SEA-LION v4.5 E2B-ITlocal | 5/683.3% | 1/2 | - | 920 msmedian | |
| Qwen3 30B A3B Instruct 2507Alibaba | 2/633.3% | 0/2 | $0.000050$0.0001 | 477 msmedian | |
| Llama3 8B CPT Sahabat-AI v1 Instructlocal | 1/616.7% | 0/2 | - | 426 msmedian | |
| Llama 3.1 8B Instructlocal | 1/616.7% | 0/2 | - | 571 msmedian | |
| Mistral Small 3Mistral | 0/60.0% | 0/2 | -$0.0001 | 1,098 msmedian |
Models marked local run on our own hardware, so there is no bill to record and the wait time measures our machine, not a service anyone else can buy. The score is still comparable: the items, the key, and the temperature are identical to every other row here.
Every figure in this table can be recomputed from the raw data. Open the data →
Cost and tokensmultiple choiceCheapest per correct answer: Solar Pro 4 ($0.000010). Most expensive: Gemini 3.5 Flash ($0.002148).Show chartsHide
All cost figures are in US dollars.
Cost per correct answer · multiple choice
The model with the highest score is not always the cheapest one. This divides the cost of one full test by its number of correct answers.
Cost of one full run · multiple choice
The raw figure, before dividing by the number of correct answers. Every model answered the exact same questions, so this is directly comparable: it is what you pay to run this benchmark once on each model.
Tokens used · multiple choice
Input and output are kept apart because they are priced differently, often tenfold. A long output bar means the model talks a lot, and that is where almost all of the cost difference between models comes from. A model with no bar has no token record, which is not the same as zero.
Cost and tokensopen answerCheapest per correct answer: Gemma 4 31B ($0.000019). Most expensive: Gemini 3.5 Flash ($0.003273).Show chartsHide
All cost figures are in US dollars.
Cost per correct answer · open answer
The model with the highest score is not always the cheapest one. This divides the cost of one full test by its number of correct answers.
Cost of one full run · open answer
The raw figure, before dividing by the number of correct answers. Every model answered the exact same questions, so this is directly comparable: it is what you pay to run this benchmark once on each model.
Tokens used · open answer
Input and output are kept apart because they are priced differently, often tenfold. A long output bar means the model talks a lot, and that is where almost all of the cost difference between models comes from. A model with no bar has no token record, which is not the same as zero.