🏆 Quality winner: openrouter:deepseek/deepseek-v4-pro
Run 2eb734 combines flawless fact-grounding with the best analogies in the field—the Tu-4/B-29 Cold War reverse-engineering example precisely mirrors distillation's mechanism (a cheaper copy that doesn't match the original but closes a strategic gap), which is exactly the article's point. It has a complete 10-question quiz with well-constructed evidence-pairing and vocab-in-context items all answerable from its original, complete, engaging passage, plus a rich six-term glossary. c2b157 and e2507b are close runners-up but slightly less distinctive in analogy; several otherwise-strong entries lost points for truncated rewritten passages.
| # | Model | Facts | Analogy | Age-fit | Clarity | Quiz | Glossary | Passage | Overall | Cost | Value | Notes |
|---|
| 🥇 |
openrouter:deepseek/deepseek-v4-pro
view ↗ |
10 | 9 | 9 | 9 | 9 | 9 | 9 |
9 |
0.62¢ |
8.7 |
Facts are accurate and richly detailed (Tu-4 analogy is apt and shares mechanism of reverse-engineering with cost gap). Full 10-question quiz with strong SAT-style evidence pairing and vocab-in-context items answerable from the rewritten passage. Original, engaging passage. Very strong all around. |
| 🥈 |
openrouter:minimax/minimax-m2.7
view ↗ |
10 | 8 | 9 | 9 | 9 | 9 | 9 |
9 |
0.62¢ |
8.7 |
Highly accurate, well-structured, with a clear F1-car and university analogy that captures the cost-vs-effort mechanism. Full 10-question quiz well-anchored to the passage; glossary precise. Passage is original and readable. |
| 🥉 |
openrouter:moonshotai/kimi-k2.6
view ↗ |
9 | 9 | 9 | 9 | 9 | 9 | 9 |
9 |
9.62¢ |
6.6 |
Excellent conservatory/pharma analogies matching the same mechanism. Rewritten passage is complete, original and sophisticated. Quiz is well-crafted with fair evidence-pairing. Faithful to source facts including McGuire and Liu Pengyu. |
| 4 |
openrouter:mistralai/mistral-large-2512
view ↗ |
9 | 8 | 9 | 9 | 8 | 8 | 9 |
8 |
0.85¢ |
8.0 |
Solid, faithful, engaging. 'Surreptitious' vocab question (Q5) is grounded in the source's word but the rewritten passage itself doesn't use it, a minor mismatch. Good analogies (cheat code, supply chain heist) and a strong original passage. |
| 5 |
openrouter:stepfun/step-3.7-flash
view ↗ |
9 | 9 | 9 | 8 | 8 | 9 | 8 |
8 |
2.09¢ |
6.8 |
Strong content and excellent SAT analogies (valedictorian, burger joint). Q5 tests 'surreptitious' which appears in source but not clearly in the truncated rewritten passage. The rewritten_passage is cut off mid-sentence, hurting completeness. |
| 6 |
openrouter:qwen/qwen3-max-thinking
view ↗ |
9 | 8 | 9 | 8 | 8 | 8 | 7 |
8 |
1.02¢ |
7.4 |
Good quiz and glossary; answer-key analogy is apt. But rewritten_passage is truncated mid-sentence, reducing quality. Q5 'surreptitious' vocab relies on a word that may not appear in the truncated passage. |
| 7 |
openrouter:x-ai/grok-4.3
view ↗ |
9 | 7 | 8 | 8 | 8 | 8 | 8 |
8 |
1.64¢ |
7.4 |
Faithful and clean; analogies (reverse-engineering, scouting playbook) are decent but somewhat generic. Complete passage, solid full quiz. Slightly less vivid than top entries but accurate throughout. |
| 8 |
openrouter:google/gemini-3.1-flash-lite
view ↗ |
8 | 8 | 9 | 9 | 9 | 8 | 8 |
8 |
0.59¢ |
8.0 |
Clean structure, chef-recipe and capture-the-flag analogies fit reasonably. Full 10-question quiz with good SAT scaffolding. Complete, engaging passage. Solid and well-rounded. |
| 9 |
openrouter:z-ai/glm-5-turbo
view ↗ |
8 | 8 | 9 | 9 | 8 | 9 | 8 |
8 |
3.40¢ |
6.8 |
Strong music-producer/scouting analogies and clear structure. Q4 tests 'sharp' ('drew a sharp distinction') which the writer's own rewritten passage uses—clever and self-consistent. Rewritten passage truncated at end, a small deduction. |
| 10 |
openrouter:openai/gpt-5.4-nano
view ↗ |
9 | 7 | 9 | 9 | 9 | 9 | 9 |
8 |
0.63¢ |
8.0 |
Careful to attribute claims to 'the article,' which is scrupulously faithful though slightly repetitive. Strong 6-term glossary, full solid quiz, complete engaging passage. Analogies (textbook/study-guide, film scouting) are decent. |
| 11 |
anthropic:claude-haiku-4-5-20251001
view ↗ |
9 | 8 | 9 | 9 | 9 | 9 | 8 |
8 |
2.57¢ |
6.8 |
Very faithful with named companies and dates. Arms-race and doping analogies work; doping is slightly generic. Full quiz with excellent evidence pairing and vocab items. Rewritten passage is complete and strong but ends slightly truncated. |
| 12 |
openrouter:amazon/nova-pro-v1
view ↗ |
8 | 6 | 7 | 7 | 7 | 7 | 7 |
7 |
1.34¢ |
6.7 |
Accurate but thinner. Chess/doping analogies are generic and doping doesn't share the distillation mechanism well. Passage is a compressed rewrite that's serviceable. Only 3 glossary terms; quiz is fine but less rigorous. |
| 13 |
openrouter:meta-llama/llama-4-maverick
view ↗ |
8 | 6 | 8 | 7 | 8 | 8 | 7 |
7 |
0.26¢ |
7.9 |
Accurate but analogies are brief and generic (student copying, reverse-engineering). Explanation sections are short. Quiz is complete and fair. Passage is competent but less vivid than leaders. |
| 14 |
openrouter:bytedance-seed/seed-2.0-lite
view ↗ |
7 | 6 | 8 | 8 | 7 | 8 | 7 |
7 |
1.73¢ |
6.7 |
Serviceable and mostly accurate but adds an unsourced '$1 trillion industry' figure. Analogies are ok. Q4 tests 'principal'/'distil' meaning that may not be perfectly grounded in the rewritten passage. Reasonable overall. |
| 15 |
openrouter:baidu/ernie-4.5-vl-424b-a47b
view ↗ |
7 | 6 | 8 | 7 | 6 | 7 | 7 |
6 |
0.43¢ |
7.2 |
Accurate but thin: only 3 quiz questions and 3 glossary terms. Analogies (chef, counterfeit) are generic. Passage is competent but compressed. Least developed of the fuller entries. |
| 16 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
Duplicate id placeholder; ignore. |
| 17 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
placeholder |
| 18 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
placeholder |
| 19 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
placeholder |
| 20 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
placeholder |
| 21 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
placeholder |
| 22 |
|
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 |
– |
– |
placeholder |
Scores 0–10 · 🟩 ≥8 · 🟨 4–7 · 🟥 <4 · sorted by Overall. Hover a column header for its full rubric name.
| # | Model | Quality | Cost | Value |
|---|
| 🥇 |
openrouter:deepseek/deepseek-v4-pro |
9 |
0.62¢ |
8.7 |
| 🥈 |
openrouter:minimax/minimax-m2.7 |
9 |
0.62¢ |
8.7 |
| 🥉 |
openrouter:mistralai/mistral-large-2512 |
8 |
0.85¢ |
8.0 |
| 4 |
openrouter:google/gemini-3.1-flash-lite |
8 |
0.59¢ |
8.0 |
| 5 |
openrouter:openai/gpt-5.4-nano |
8 |
0.63¢ |
8.0 |
| 6 |
openrouter:meta-llama/llama-4-maverick |
7 |
0.26¢ |
7.9 |
| 7 |
openrouter:qwen/qwen3-max-thinking |
8 |
1.02¢ |
7.4 |
| 8 |
openrouter:x-ai/grok-4.3 |
8 |
1.64¢ |
7.4 |
| 9 |
openrouter:baidu/ernie-4.5-vl-424b-a47b |
6 |
0.43¢ |
7.2 |
| 10 |
openrouter:stepfun/step-3.7-flash |
8 |
2.09¢ |
6.8 |
| 11 |
openrouter:z-ai/glm-5-turbo |
8 |
3.40¢ |
6.8 |
| 12 |
anthropic:claude-haiku-4-5-20251001 |
8 |
2.57¢ |
6.8 |
| 13 |
openrouter:amazon/nova-pro-v1 |
7 |
1.34¢ |
6.7 |
| 14 |
openrouter:bytedance-seed/seed-2.0-lite |
7 |
1.73¢ |
6.7 |
| 15 |
openrouter:moonshotai/kimi-k2.6 |
9 |
9.62¢ |
6.6 |
| # | Model | Quality | Cost | Value |
|---|
| 1 |
openrouter:deepseek/deepseek-v4-pro |
9 |
0.62¢ |
8.7 |
| 2 |
openrouter:minimax/minimax-m2.7 |
9 |
0.62¢ |
8.7 |
| 3 |
openrouter:mistralai/mistral-large-2512 |
8 |
0.85¢ |
8.0 |
| 4 |
openrouter:google/gemini-3.1-flash-lite |
8 |
0.59¢ |
8.0 |
| 5 |
openrouter:openai/gpt-5.4-nano |
8 |
0.63¢ |
8.0 |
| 6 |
openrouter:qwen/qwen3-max-thinking |
8 |
1.02¢ |
7.4 |
| 7 |
openrouter:x-ai/grok-4.3 |
8 |
1.64¢ |
7.4 |
| 8 |
openrouter:stepfun/step-3.7-flash |
8 |
2.09¢ |
6.8 |
| 9 |
anthropic:claude-haiku-4-5-20251001 |
8 |
2.57¢ |
6.8 |
| 10 |
openrouter:meta-llama/llama-4-maverick |
7 |
0.26¢ |
7.9 |
| 11 |
openrouter:amazon/nova-pro-v1 |
7 |
1.34¢ |
6.7 |
| 12 |
openrouter:bytedance-seed/seed-2.0-lite |
7 |
1.73¢ |
6.7 |
| 13 |
openrouter:baidu/ernie-4.5-vl-424b-a47b |
6 |
0.43¢ |
7.2 |
Strong quality without paying flagship prices — the cheap-and-good picks.