Topics
The 2,198 questions fall into 20 topics of the book: 6 test reading and 14 test grammar. Each topic's page lists its questions and every setup's result on it.
- Word and phrase meanings Reading · 131 questions
- Sentence meanings Reading · 154 questions
- How a writer explains and develops ideas Reading · 70 questions
- Topic and main point of a paragraph Reading · 98 questions
- How a paragraph fits together Reading · 106 questions
- Details that support a paragraph’s main point Reading · 97 questions
- Word types, such as nouns and verbs Grammar · 140 questions
- Groups of words built around a noun Grammar · 65 questions
- Verbs and their forms Grammar · 91 questions
- Verb forms used like nouns or modifiers Grammar · 72 questions
- Word endings and what they do Grammar · 79 questions
- How words are built Grammar · 69 questions
- The roles of words and phrases in a sentence Grammar · 145 questions
- How verbs show who acts and who receives the action Grammar · 74 questions
- Types of sentences Grammar · 85 questions
- Questions combining several grammar topics Grammar · 220 questions
- Sound patterns and changes Grammar · 77 questions
- Spelling Grammar · 133 questions
- Punctuation Grammar · 117 questions
- Problems in sentence wording Grammar · 175 questions
Accuracy by topic
Each cell: the share of the topic's questions the setup got right, coloured from dark blue at 0% through turquoise and green to yellow at 100%. Setups by accuracy on all questions; the table scrolls sideways inside its box.
| Setup | Word and phrase meaningsReading · 131 | Sentence meaningsReading · 154 | How a writer explains and develops ideasReading · 70 | Topic and main point of a paragraphReading · 98 | How a paragraph fits togetherReading · 106 | Details that support a paragraph’s main pointReading · 97 | Word types, such as nouns and verbsGrammar · 140 | Groups of words built around a nounGrammar · 65 | Verbs and their formsGrammar · 91 | Verb forms used like nouns or modifiersGrammar · 72 | Word endings and what they doGrammar · 79 | How words are builtGrammar · 69 | The roles of words and phrases in a sentenceGrammar · 145 | How verbs show who acts and who receives the actionGrammar · 74 | Types of sentencesGrammar · 85 | Questions combining several grammar topicsGrammar · 220 | Sound patterns and changesGrammar · 77 | SpellingGrammar · 133 | PunctuationGrammar · 117 | Problems in sentence wordingGrammar · 175 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 3.8 Flashhigh reasoning · 16 questions per request | 97% | 99% | 94% | 98% | 93% | 99% | 99% | 100% | 100% | 99% | 99% | 97% | 98% | 99% | 99% | 99% | 100% | 100% | 98% | 96% |
| GPT-6 Astrahigh reasoning · 16 questions per request | 98% | 99% | 93% | 97% | 95% | 99% | 99% | 98% | 100% | 99% | 99% | 96% | 98% | 99% | 99% | 97% | 99% | 99% | 97% | 98% |
| GPT-6.1 Solhigh reasoning · 16 questions per request | 98% | 98% | 91% | 96% | 95% | 98% | 99% | 98% | 100% | 99% | 99% | 96% | 98% | 99% | 99% | 99% | 99% | 99% | 96% | 98% |
| GPT-6 Astralow reasoning · 16 questions per request | 98% | 99% | 93% | 97% | 94% | 98% | 99% | 98% | 100% | 99% | 99% | 94% | 98% | 99% | 99% | 96% | 96% | 99% | 97% | 98% |
| GPT-6 Astralow reasoning · 1 question per request | 98% | 99% | 94% | 98% | 95% | 98% | 98% | 97% | 100% | 97% | 99% | 96% | 97% | 99% | 99% | 96% | 97% | 100% | 93% | 98% |
| Gemini 3.8 Flashlow reasoning · 1 question per request | 96% | 99% | 91% | 98% | 92% | 98% | 98% | 98% | 100% | 99% | 96% | 96% | 98% | 100% | 98% | 98% | 96% | 98% | 96% | 97% |
| GPT-6.1 Sollow reasoning · 1 question per request | 97% | 97% | 91% | 96% | 93% | 99% | 98% | 98% | 98% | 99% | 99% | 94% | 97% | 99% | 99% | 95% | 94% | 98% | 94% | 98% |
| GPT-6.1 Sollow reasoning · 16 questions per request | 98% | 98% | 91% | 96% | 93% | 98% | 99% | 97% | 100% | 99% | 99% | 94% | 98% | 99% | 96% | 95% | 94% | 98% | 94% | 98% |
| Gemini 4 Argonlow reasoning · 16 questions per request | 97% | 100% | 90% | 99% | 96% | 96% | 96% | 98% | 100% | 99% | 95% | 96% | 96% | 99% | 96% | 92% | 95% | 98% | 96% | 96% |
| Claude Fable 5.1high reasoning · 16 questions per request | 95% | 97% | 91% | 98% | 92% | 98% | 96% | 97% | 100% | 99% | 99% | 96% | 96% | 100% | 99% | 94% | 95% | 96% | 92% | 97% |
| Claude Opus 5.5high reasoning · 16 questions per request | 96% | 98% | 90% | 99% | 91% | 97% | 97% | 98% | 98% | 99% | 99% | 94% | 97% | 96% | 95% | 94% | 91% | 90% | 89% | 94% |
| GPT-6.1 Terrahigh reasoning · 16 questions per request | 95% | 99% | 91% | 97% | 95% | 97% | 97% | 95% | 98% | 99% | 95% | 91% | 95% | 95% | 98% | 88% | 87% | 95% | 85% | 95% |
| Claude Opus 5.5low reasoning · 1 question per request | 98% | 97% | 86% | 97% | 94% | 99% | 96% | 97% | 98% | 99% | 97% | 93% | 96% | 95% | 92% | 90% | 86% | 85% | 82% | 94% |
| Gemini 3.8 Flashlow reasoning · 16 questions per request | 96% | 97% | 89% | 98% | 91% | 98% | 95% | 95% | 100% | 94% | 91% | 93% | 90% | 99% | 93% | 86% | 86% | 96% | 89% | 95% |
| Grok 4.7high reasoning · 16 questions per request | 95% | 99% | 86% | 96% | 92% | 97% | 94% | 97% | 96% | 100% | 94% | 88% | 97% | 97% | 95% | 85% | 90% | 86% | 85% | 93% |
| DeepSeek-V4.1-Flash, first releasereasoning on · 16 questions per request | 95% | 97% | 84% | 94% | 89% | 96% | 92% | 97% | 98% | 97% | 95% | 88% | 94% | 95% | 95% | 86% | 86% | 94% | 84% | 90% |
| GPT-6.1 Terralow reasoning · 16 questions per request | 97% | 97% | 91% | 96% | 92% | 99% | 94% | 89% | 96% | 96% | 92% | 78% | 90% | 91% | 94% | 82% | 71% | 90% | 84% | 95% |
| Claude Opus 5.5low reasoning · 16 questions per request | 95% | 97% | 86% | 98% | 95% | 94% | 98% | 98% | 95% | 97% | 95% | 84% | 96% | 96% | 93% | 74% | 70% | 83% | 85% | 91% |
| Claude Sonnet 5.5high reasoning · 1 question per request | 95% | 96% | 83% | 96% | 95% | 96% | 94% | 94% | 95% | 93% | 95% | 88% | 90% | 88% | 94% | 84% | 74% | 77% | 76% | 88% |
| Claude Sonnet 5.5high reasoning · 16 questions per request | 80% | 90% | 69% | 93% | 82% | 93% | 94% | 91% | 87% | 94% | 95% | 84% | 90% | 92% | 91% | 77% | 79% | 76% | 80% | 79% |
| Kimi K3.1 Prolow reasoning · 16 questions per request | 92% | 99% | 81% | 98% | 91% | 97% | 83% | 85% | 87% | 93% | 81% | 74% | 80% | 84% | 81% | 67% | 52% | 77% | 72% | 86% |
| GLM-5.3-Flashreasoning on · 16 questions per request | 89% | 92% | 79% | 96% | 82% | 95% | 84% | 85% | 87% | 90% | 89% | 77% | 84% | 89% | 85% | 63% | 0% | 65% | 70% | 81% |
| Liquid d1-ultra1 question per request | 89% | 96% | 86% | 96% | 91% | 93% | 79% | 77% | 82% | 76% | 68% | 46% | 72% | 77% | 75% | 62% | 39% | 68% | 75% | 82% |
| GLM-5.3low reasoning · 16 questions per request | 85% | 94% | 74% | 96% | 81% | 91% | 78% | 65% | 76% | 90% | 78% | 55% | 73% | 70% | 75% | 63% | 34% | 67% | 62% | 74% |
| Qwen3.8-27Blow reasoning · 16 questions per request | 86% | 96% | 79% | 96% | 81% | 93% | 78% | 80% | 80% | 85% | 80% | 38% | 79% | 80% | 78% | 44% | 26% | 65% | 66% | 77% |
| Gemma 4 31Breasoning off · 16 questions per request | 86% | 94% | 89% | 97% | 84% | 96% | 67% | 65% | 76% | 79% | 68% | 55% | 64% | 77% | 68% | 50% | 49% | 74% | 70% | 70% |
| Kimi K3.1 Proreasoning none · 16 questions per request | 91% | 96% | 83% | 98% | 86% | 98% | 67% | 55% | 76% | 71% | 65% | 51% | 61% | 82% | 68% | 50% | 31% | 73% | 65% | 79% |
| Fastino GLiDE v1.2local · 1 question per request | 89% | 94% | 86% | 93% | 78% | 96% | 64% | 60% | 62% | 67% | 66% | 41% | 58% | 70% | 62% | 47% | 35% | 60% | 65% | 73% |
| GPT-6 Lunalow reasoning · 16 questions per request | 87% | 94% | 74% | 95% | 75% | 94% | 55% | 48% | 70% | 68% | 65% | 52% | 54% | 74% | 65% | 41% | 38% | 82% | 65% | 73% |
| Fastino GLiDE1 question per request | 82% | 94% | 83% | 91% | 83% | 92% | 62% | 62% | 65% | 69% | 62% | 45% | 68% | 69% | 64% | 45% | 32% | 60% | 55% | 73% |
| vLLM-SR d31 question per request | 85% | 92% | 81% | 96% | 84% | 94% | 66% | 60% | 63% | 71% | 62% | 42% | 59% | 68% | 62% | 45% | 31% | 59% | 62% | 71% |
| vLLM-SR d3-ultratransformers, rev 52ba9e7 · 1 question per request | 85% | 93% | 83% | 95% | 86% | 95% | 65% | 54% | 60% | 68% | 62% | 39% | 59% | 76% | 69% | 45% | 27% | 61% | 62% | 70% |
| Drex Max v11 question per request | 84% | 93% | 81% | 94% | 88% | 96% | 66% | 60% | 64% | 69% | 65% | 41% | 52% | 68% | 65% | 46% | 32% | 59% | 62% | 71% |
| Drex Max 1.11 question per request | 85% | 94% | 84% | 94% | 81% | 92% | 66% | 57% | 63% | 68% | 65% | 43% | 54% | 61% | 66% | 46% | 34% | 62% | 62% | 70% |
| Mammoth-11 question per request | 85% | 92% | 83% | 96% | 87% | 99% | 69% | 58% | 58% | 58% | 65% | 49% | 52% | 70% | 52% | 40% | 39% | 58% | 64% | 71% |
| Perplexity Decider 27B v1.2local · 1 question per request | 83% | 94% | 84% | 94% | 85% | 94% | 66% | 58% | 60% | 65% | 62% | 41% | 54% | 66% | 60% | 44% | 27% | 61% | 63% | 73% |
| Perplexity Decider 27B v1.11 question per request | 84% | 93% | 84% | 95% | 83% | 94% | 64% | 58% | 65% | 65% | 63% | 39% | 53% | 72% | 61% | 42% | 25% | 62% | 62% | 72% |
| Gemma 4 31Breasoning off · 1 question per request | 86% | 92% | 87% | 96% | 79% | 94% | 65% | 52% | 66% | 62% | 61% | 48% | 48% | 65% | 51% | 40% | 42% | 70% | 63% | 71% |
| DeepSeek-V4.1-Flash, first releasereasoning off · 16 questions per request | 85% | 91% | 80% | 96% | 80% | 93% | 60% | 52% | 66% | 64% | 58% | 41% | 48% | 72% | 54% | 37% | 40% | 77% | 56% | 73% |
| SubQ Snap Preview1 question per request | 81% | 95% | 76% | 92% | 82% | 96% | 61% | 54% | 57% | 68% | 61% | 36% | 54% | 68% | 56% | 43% | 26% | 56% | 61% | 73% |
| SubQ Snap 1.11 question per request | 82% | 92% | 79% | 92% | 80% | 94% | 64% | 55% | 55% | 64% | 59% | 35% | 50% | 59% | 56% | 41% | 19% | 59% | 60% | 69% |
| GPT-6 LunaDecisions API, beta · 1 question per request | 87% | 88% | 73% | 95% | 75% | 91% | 50% | 42% | 66% | 62% | 48% | 45% | 51% | 65% | 58% | 36% | 43% | 76% | 53% | 68% |
| Liquid d1.11 question per request | 77% | 92% | 76% | 95% | 75% | 90% | 50% | 48% | 58% | 64% | 52% | 33% | 57% | 58% | 51% | 41% | 32% | 50% | 60% | 64% |
| Claude Sonnet 5.5low reasoning · 16 questions per request | 82% | 86% | 71% | 93% | 72% | 90% | 54% | 48% | 67% | 61% | 53% | 41% | 44% | 72% | 48% | 31% | 35% | 70% | 53% | 66% |
| Celeris-1 Decision1 question per request | 79% | 91% | 81% | 89% | 78% | 91% | 55% | 52% | 62% | 50% | 52% | 38% | 46% | 55% | 56% | 35% | 42% | 45% | 51% | 66% |
| Perplexity Decider 27B v11 question per request | 79% | 89% | 80% | 94% | 78% | 96% | 54% | 51% | 60% | 46% | 57% | 29% | 44% | 66% | 58% | 36% | 31% | 50% | 50% | 62% |
| OneJev 27B1 question per request | 81% | 86% | 80% | 93% | 78% | 93% | 51% | 48% | 58% | 51% | 59% | 33% | 44% | 58% | 52% | 35% | 32% | 50% | 52% | 62% |
| Jev 1.14.01 question per request | 79% | 90% | 74% | 95% | 77% | 92% | 56% | 48% | 58% | 53% | 48% | 35% | 46% | 58% | 51% | 37% | 34% | 41% | 51% | 61% |
| Inception Mercury Decidevia OpenRouter · 1 question per request | 79% | 89% | 74% | 94% | 76% | 92% | 56% | 46% | 51% | 43% | 52% | 32% | 39% | 59% | 46% | 35% | 35% | 51% | 50% | 62% |
| GLM-5.3-Flashreasoning off · 16 questions per request | 76% | 84% | 63% | 93% | 78% | 94% | 52% | 42% | 54% | 49% | 48% | 30% | 39% | 59% | 45% | 34% | 36% | 57% | 50% | 65% |
| Jev 1.13.01 question per request | 80% | 89% | 71% | 95% | 74% | 89% | 54% | 49% | 56% | 57% | 49% | 35% | 45% | 59% | 48% | 31% | 27% | 39% | 47% | 61% |
| Clef1 question per request | 77% | 92% | 77% | 95% | 75% | 91% | 51% | 42% | 52% | 42% | 52% | 36% | 34% | 51% | 49% | 34% | 31% | 53% | 47% | 61% |
| Liquid d11 question per request | 79% | 86% | 76% | 95% | 72% | 88% | 49% | 51% | 51% | 56% | 49% | 35% | 46% | 54% | 47% | 29% | 29% | 46% | 49% | 63% |
| Drex 2.01 question per request | 73% | 84% | 76% | 94% | 75% | 87% | 49% | 40% | 57% | 49% | 42% | 36% | 42% | 61% | 42% | 33% | 26% | 45% | 50% | 57% |
| Cygnet1 question per request | 76% | 91% | 69% | 96% | 72% | 88% | 50% | 42% | 59% | 51% | 49% | 36% | 37% | 51% | 40% | 29% | 31% | 46% | 46% | 58% |
| Open-Jev 27B v1.11 question per request | 77% | 85% | 71% | 94% | 75% | 93% | 51% | 52% | 57% | 46% | 59% | 28% | 40% | 54% | 41% | 18% | 23% | 52% | 47% | 62% |
| Winnow-12B1 question per request | 76% | 90% | 67% | 93% | 75% | 87% | 46% | 43% | 51% | 43% | 48% | 38% | 40% | 43% | 39% | 34% | 34% | 47% | 49% | 60% |
| Qwen3.8-27Bvote share · reasoning off · 1 question per request | 81% | 86% | 74% | 93% | 77% | 90% | 42% | 34% | 46% | 38% | 42% | 36% | 41% | 45% | 44% | 24% | 31% | 50% | 50% | 59% |
| Qwen3.8-27Bstated confidence · reasoning off · 1 question per request | 78% | 86% | 74% | 93% | 75% | 93% | 39% | 34% | 45% | 39% | 49% | 28% | 43% | 46% | 41% | 24% | 38% | 48% | 50% | 60% |
| Elephant-11 question per request | 73% | 87% | 69% | 94% | 71% | 86% | 46% | 38% | 36% | 39% | 43% | 29% | 39% | 58% | 32% | 30% | 21% | 38% | 47% | 59% |
| Qwen3.8-27Byes/no scoring · 1 question per request | 69% | 88% | 73% | 91% | 75% | 86% | 42% | 29% | 45% | 38% | 37% | 30% | 38% | 45% | 32% | 30% | 32% | 47% | 40% | 58% |
| Perplexity Decider Flash v1.2local · 1 question per request | 62% | 80% | 70% | 87% | 58% | 80% | 45% | 38% | 41% | 53% | 42% | 26% | 33% | 38% | 36% | 27% | 16% | 32% | 39% | 53% |
| Microsoft-Decision-1via OpenRouter · 1 question per request | 66% | 79% | 70% | 88% | 63% | 75% | 44% | 29% | 44% | 33% | 37% | 30% | 32% | 41% | 29% | 25% | 23% | 33% | 36% | 47% |
| Clef-flash1 question per request | 59% | 86% | 60% | 88% | 61% | 79% | 37% | 38% | 36% | 35% | 38% | 25% | 32% | 35% | 31% | 26% | 29% | 30% | 38% | 46% |
| Qwen3-14Breasoning off · 16 questions per request | 61% | 75% | 66% | 86% | 68% | 78% | 32% | 29% | 30% | 33% | 30% | 36% | 26% | 34% | 32% | 28% | 31% | 38% | 45% | 52% |
| Erk-14Breasoning off · 16 questions per request | 64% | 74% | 66% | 91% | 60% | 79% | 32% | 31% | 30% | 39% | 38% | 23% | 26% | 28% | 29% | 27% | 26% | 36% | 34% | 55% |
| Metask-Jev 4B1 question per request | 50% | 69% | 54% | 86% | 49% | 68% | 34% | 34% | 40% | 43% | 33% | 30% | 28% | 26% | 25% | 25% | 34% | 27% | 27% | 41% |
| Liquid d1-2B (open weights)1 question per request | 38% | 44% | 39% | 55% | 28% | 41% | 20% | 23% | 27% | 21% | 25% | 22% | 27% | 34% | 26% | 21% | 25% | 20% | 22% | 32% |
| Strands Decider 2B1 question per request | 32% | 46% | 33% | 60% | 33% | 12% | 25% | 23% | 32% | 24% | 24% | 14% | 18% | 20% | 20% | 22% | 19% | 26% | 25% | 29% |
| meraGPT Decider 11 question per request | 27% | 30% | 34% | 54% | 31% | 37% | 30% | 26% | 25% | 19% | 29% | 20% | 25% | 22% | 15% | 17% | 22% | 21% | 26% | 33% |
| GLiNER2.5-multi-Decide1 question per request | 22% | 23% | 17% | 22% | 14% | 25% | 18% | 12% | 25% | 12% | 18% | 20% | 26% | 24% | 18% | 18% | 19% | 20% | 23% | 21% |
| Laya1 question per request | 25% | 23% | 11% | 18% | 26% | 15% | 13% | 15% | 16% | 14% | 16% | 14% | 23% | 16% | 27% | 21% | 22% | 17% | 15% | 19% |
| Laya multilingual1 question per request | 19% | 12% | 16% | 21% | 19% | 16% | 19% | 12% | 15% | 17% | 20% | 16% | 17% | 15% | 20% | 19% | 17% | 22% | 17% | 22% |
| GLiNER2.5 Multi (base)1 question per request | 18% | 13% | 17% | 21% | 21% | 14% | 16% | 14% | 18% | 15% | 22% | 13% | 19% | 20% | 22% | 15% | 17% | 14% | 15% | 26% |