Relationship tree
Extracts a deterministic XML family and relationship tree from cumulative literary prose.
Performance Score Distribution (Top 20)
Click a model name to view its detail page.
Price-Performance Score Distribution (Top 20)
Click a model name to view its detail page.
| Score | Cost | Time | ||
|---|---|---|---|---|
| Gemini 4 Argon (Reasoning, Medium) | 95% | $0.039 | 19.1s | |
| Gemini 3.7 Flash (Reasoning, Medium) | 94% | $0.036 | 23.4s | |
| GPT-6.1 Sol (Reasoning, Medium) | 97% | $0.042 | 50.1s | |
| GPT-6 Sol (Reasoning, Medium) | 96% | $0.051 | 46.3s | |
| Gemini 3.8 Flash (Reasoning, Medium) | 96% | $0.067 | 42.1s | |
| Claude Sonnet 5.5 (Reasoning) | 94% | $0.101 | 28.5s | |
| GPT-6 Luna (Reasoning, High) | 94% | $0.0055 | 1.4m | |
| DeepSeek V4.1 Flash 1017 (Reasoning, High) | 92% | $0.0056 | 58.2s | |
| GPT-6 Sol | 90% | $0.035 | 21.1s | |
| Muse Spark 1.2 (Reasoning, Medium) | 90% | $0.046 | 29.8s | |
| Muse Wonder (Reasoning, Medium) | 94% | $0.116 | 20.5s | |
| Gemini 3.6 Flash (Reasoning) | 94% | $0.096 | 41.4s | |
| Claude Haiku 5.5 (Adaptive) | 88% | $0.0064 | 29.8s | |
| GPT-5.6 Terra (Reasoning, Medium) | 91% | $0.060 | 24.2s | |
| GPT-5.5 (Reasoning, Low) | 87% | $0.097 | 42.6s | |
| GPT-6 Luna (Reasoning, Medium) | 90% | $0.0035 | 1.1m | |
| Xiaomi MIMO v2.6 Pro | 88% | $0.0071 | 2.0m | |
| DeepSeek V4.1 Flash (Reasoning, High) | 92% | $0.014 | 1.5m | |
| Claude Haiku 5.5 (Reasoning, Low) | 84% | $0.0050 | 19.9s | |
| DeepSeek V4 Flash 0731 (Reasoning, High) | 93% | $0.0073 | 5.6m | |
Cost vs Performance
Compares total cost for this test against the test score. Quadrant lines are drawn at the median values. Only models with available cost data are shown.
1 low-scoring outlier hidden: Gemma 3 12B (37.1%).
Most Stable Models (Top 20)
Ranked by stability (median × consistency). Click a model name to view its detail page.
| Score | Consistency | Stability | ||
|---|---|---|---|---|
| Gemini 3.8 Flash (Reasoning, Medium) | 96% | 95% | 93% | |
| Ember-1 (Reasoning, Max) | 96% | 95% | 92% | |
| Claude Opus 5.5 (Reasoning) | 96% | 95% | 91% | |
| Gemini 4 Argon (Reasoning, Medium) | 95% | 96% | 91% | |
| GPT-6 Sol (Reasoning, Medium) | 96% | 95% | 91% | |
| GPT-6.1 Sol (Reasoning, Medium) | 97% | 94% | 91% | |
| GPT-5.4 (Reasoning) | 96% | 93% | 90% | |
| Gemini 3.7 Flash (Reasoning, Medium) | 94% | 95% | 90% | |
| GPT-5.6 Sol (Reasoning, Medium) | 96% | 94% | 90% | |
| Grok 4.6 (Reasoning, High) | 94% | 94% | 89% | |
| Gemini 3.6 Flash (Reasoning) | 94% | 93% | 88% | |
| Claude Opus 4.6 (Reasoning) | 95% | 93% | 88% | |
| Claude Sonnet 5.5 (Reasoning) | 94% | 91% | 88% | |
| Z.AI GLM 5.3 (Reasoning, Max) | 94% | 90% | 88% | |
| GPT-5.5 (Reasoning, Medium) | 94% | 93% | 87% | |
| Grok 4.7 (Reasoning, High) | 96% | 92% | 87% | |
| GPT-6 Luna (Reasoning, High) | 94% | 93% | 87% | |
| Muse Wonder (Reasoning, Medium) | 94% | 93% | 87% | |
| Claude Opus 4.8 (Reasoning) | 94% | 90% | 86% | |
| Qwen 3.8 Flash (Reasoning) | 94% | 90% | 86% | |
Top Overall Models (Top 20)
Ranked by composite score (performance, cost, speed & stability). Click a model name to view its detail page.
| Score | Cost | Speed | Stability | ||
|---|---|---|---|---|---|
| Gemini 4 Argon (Reasoning, Medium) | 95% | $0.039 | 19.1s | 91% | |
| Gemini 3.7 Flash (Reasoning, Medium) | 94% | $0.036 | 23.4s | 90% | |
| GPT-6.1 Sol (Reasoning, Medium) | 97% | $0.042 | 50.1s | 91% | |
| Gemini 3.8 Flash (Reasoning, Medium) | 96% | $0.067 | 42.1s | 93% | |
| GPT-6 Sol (Reasoning, Medium) | 96% | $0.051 | 46.3s | 91% | |
| GPT-6 Luna (Reasoning, High) | 94% | $0.0055 | 1.4m | 87% | |
| DeepSeek V4.1 Flash 1017 (Reasoning, High) | 92% | $0.0056 | 58.2s | 82% | |
| Claude Sonnet 5.5 (Reasoning) | 94% | $0.101 | 28.5s | 88% | |
| Gemini 3.6 Flash (Reasoning) | 94% | $0.096 | 41.4s | 88% | |
| GPT-6 Sol | 90% | $0.035 | 21.1s | 81% | |
| DeepSeek V4.1 Flash (Reasoning, High) | 92% | $0.014 | 1.5m | 82% | |
| GPT-6 Luna (Reasoning, Medium) | 90% | $0.0035 | 1.1m | 80% | |
| Muse Wonder (Reasoning, Medium) | 94% | $0.116 | 20.5s | 87% | |
| Claude Haiku 5.5 (Adaptive) | 88% | $0.0064 | 29.8s | 78% | |
| Claude Haiku 5.5 (Reasoning, Low) | 84% | $0.0050 | 19.9s | 78% | |
| GPT-5.6 Terra (Reasoning, Medium) | 91% | $0.060 | 24.2s | 79% | |
| Muse Spark 1.2 (Reasoning, Medium) | 90% | $0.046 | 29.8s | 77% | |
| Muse Spark 1.1 (Reasoning, Medium) | 90% | $0.042 | 35.1s | 77% | |
| GPT-5.4 (Reasoning, Low) | 88% | $0.053 | 35.5s | 80% | |
| Claude Opus 5.5 (Reasoning) | 96% | $0.203 | 40.6s | 91% | |
| Model | Total ▼ | Core relationship tree | Family relationship tree |
|---|---|---|---|
| GPT-6.1 Sol (Reasoning, Medium) | 97% | 100% | 94% |
| Gemini 3.8 Flash (Reasoning, Medium) | 96% | 98% | 94% |
| GPT-5.6 Sol (Reasoning, Medium) | 96% | 99% | 93% |
| Ember-1 (Reasoning, Max) | 96% | 98% | 94% |
| Claude Opus 5.5 (Reasoning) | 96% | 98% | 93% |
| GPT-6 Sol (Reasoning, Medium) | 96% | 98% | 93% |
| GPT-5.4 (Reasoning) | 96% | 99% | 93% |
| Grok 4.7 (Reasoning, High) | 96% | 99% | 92% |
| Claude Opus 4.6 (Reasoning) | 95% | 98% | 92% |
| Gemini 4 Argon (Reasoning, Medium) | 95% | 97% | 93% |
| Gemini 3.7 Flash (Reasoning, Medium) | 94% | 97% | 92% |
| Qwen 3.8 Flash (Reasoning) | 94% | 98% | 91% |
| Claude Sonnet 5.5 (Reasoning) | 94% | 98% | 90% |
| Z.AI GLM 5.3 (Reasoning, Max) | 94% | 98% | 90% |
| Grok 4.6 (Reasoning, High) | 94% | 97% | 91% |
1–15 of 203
Page 1 / 14
Core relationship tree
Performance Score Distribution (Top 20)
Click a model name to view its detail page.
Price-Performance Score Distribution (Top 20)
Click a model name to view its detail page.
| Score | Cost | Time | ||
|---|---|---|---|---|
| DeepSeek V4.1 Flash 1017 (Reasoning, High) | 96% | $0.0030 | 35.8s | |
| GPT-6 Luna (Reasoning, High) | 97% | $0.0028 | 44.8s | |
| GPT-6.1 Sol (Reasoning, Medium) | 100% | $0.020 | 23.7s | |
| DeepSeek V4.1 Flash (Reasoning, High) | 96% | $0.0082 | 55.5s | |
| GPT-6 Luna (Reasoning, Medium) | 92% | $0.0019 | 37.8s | |
| Gemini 4 Argon (Reasoning, Medium) | 97% | $0.021 | 12.2s | |
| GPT-6 Sol | 93% | $0.019 | 12.1s | |
| Muse Spark 1.1 (Reasoning, Minimal) | 90% | $0.022 | 15.7s | |
| Claude Haiku 5.5 (Adaptive) | 91% | $0.0038 | 18.8s | |
| GPT-6 Sol (Reasoning, Medium) | 98% | $0.028 | 27.5s | |
| Gemini 3 Flash (Preview, Reasoning) | 90% | $0.024 | 27.7s | |
| Gemini 3.7 Flash (Reasoning, Medium) | 97% | $0.024 | 16.0s | |
| Muse Spark 1.2 (Reasoning, Medium) | 98% | $0.034 | 27.3s | |
| Gemini 3.6 Flash (Reasoning, Minimal) | 84% | $0.023 | 4.9s | |
| Muse Spark 1.1 (Reasoning, Medium) | 94% | $0.030 | 26.9s | |
| DeepSeek V4 Flash 0731 (Reasoning, High) | 97% | $0.0065 | 3.1m | |
| Claude Haiku 5.5 (Reasoning, Low) | 87% | $0.0033 | 15.1s | |
| Xiaomi MIMO v2.6 Pro | 91% | $0.0038 | 58.7s | |
| GPT-5.6 Terra (Reasoning, Medium) | 95% | $0.038 | 16.3s | |
| Grok 4.5 (Reasoning, Low) | 96% | $0.033 | 50.7s | |
Most Stable Models (Top 20)
Ranked by stability (median × consistency). Click a model name to view its detail page.
| Score | Consistency | Stability | ||
|---|---|---|---|---|
| GPT-6.1 Sol (Reasoning, Medium) | 100% | 100% | 100% | |
| GPT-5.6 Sol (Reasoning, Medium) | 99% | 98% | 98% | |
| Claude Opus 5.5 (Reasoning) | 98% | 100% | 98% | |
| Gemini 3.8 Flash (Reasoning, Medium) | 98% | 100% | 98% | |
| Z.AI GLM 5.3 (Reasoning, Max) | 98% | 100% | 98% | |
| Claude Sonnet 4.6 (Reasoning) | 98% | 100% | 98% | |
| Claude Sonnet 5.5 (Reasoning) | 98% | 100% | 98% | |
| Ember-1 (Reasoning, Max) | 98% | 100% | 98% | |
| Grok 4.7 (Reasoning, High) | 99% | 97% | 97% | |
| Claude Opus 4.8 (Reasoning) | 99% | 99% | 97% | |
| Muse Spark 1.2 (Reasoning, Medium) | 98% | 98% | 96% | |
| Claude Opus 5 (Reasoning, Low) | 98% | 98% | 96% | |
| GPT-5.5 (Reasoning, Medium) | 98% | 98% | 96% | |
| GPT-6 Sol (Reasoning, Medium) | 98% | 98% | 96% | |
| Qwen 3.8 Flash (Reasoning) | 98% | 98% | 96% | |
| GPT-5.5 (Reasoning, Low) | 97% | 98% | 96% | |
| Claude Opus 4.6 (Reasoning) | 98% | 97% | 96% | |
| GPT-5.4 (Reasoning) | 99% | 98% | 96% | |
| Grok 4.5 (Reasoning, High) | 97% | 98% | 95% | |
| Gemini 3.1 Pro (Preview) | 97% | 98% | 95% | |
Top Overall Models (Top 20)
Ranked by composite score (performance, cost, speed & stability). Click a model name to view its detail page.
| Score | Cost | Speed | Stability | ||
|---|---|---|---|---|---|
| GPT-6.1 Sol (Reasoning, Medium) | 100% | $0.020 | 23.7s | 100% | |
| Gemini 4 Argon (Reasoning, Medium) | 97% | $0.021 | 12.2s | 95% | |
| GPT-6 Sol (Reasoning, Medium) | 98% | $0.028 | 27.5s | 96% | |
| Gemini 3.7 Flash (Reasoning, Medium) | 97% | $0.024 | 16.0s | 95% | |
| GPT-6 Luna (Reasoning, High) | 97% | $0.0028 | 44.8s | 92% | |
| Muse Spark 1.2 (Reasoning, Medium) | 98% | $0.034 | 27.3s | 96% | |
| Gemini 3.8 Flash (Reasoning, Medium) | 98% | $0.046 | 30.7s | 98% | |
| Claude Sonnet 5.5 (Reasoning) | 98% | $0.055 | 15.1s | 98% | |
| DeepSeek V4.1 Flash 1017 (Reasoning, High) | 96% | $0.0030 | 35.8s | 88% | |
| DeepSeek V4.1 Flash (Reasoning, High) | 96% | $0.0082 | 55.5s | 89% | |
| Claude Haiku 5.5 (Adaptive) | 91% | $0.0038 | 18.8s | 84% | |
| GPT-6 Sol | 93% | $0.019 | 12.1s | 85% | |
| GPT-6 Luna (Reasoning, Medium) | 92% | $0.0019 | 37.8s | 84% | |
| Muse Spark 1.1 (Reasoning, Medium) | 94% | $0.030 | 26.9s | 87% | |
| GPT-5.5 (Reasoning, Low) | 97% | $0.070 | 24.6s | 96% | |
| Grok 4.5 (Reasoning, Low) | 96% | $0.033 | 50.7s | 90% | |
| GPT-5.6 Terra (Reasoning, Medium) | 95% | $0.038 | 16.3s | 87% | |
| Claude Haiku 5.5 (Reasoning, Low) | 87% | $0.0033 | 15.1s | 84% | |
| GPT-5.6 Sol | 95% | $0.048 | 12.1s | 89% | |
| GPT-5.5 | 94% | $0.062 | 16.3s | 91% | |
| Median | Evaluator | Top 3 | Flop 3 |
|---|---|---|---|
| 85.4% | Alias accuracy | ||
| 100.0% | Character precision | ||
| 100.0% | Character recall | ||
| 100.0% | Isolated character handling | ||
| 100.0% | Red-herring resistance | ||
| 37.4% | Relationship category recall | ||
| 100.0% | Relationship endpoint integrity | ||
| 84.8% | Relationship precision | ||
| 32.6% | Relationship recall | ||
| 81.2% | Relationship type accuracy | ||
| 100.0% | XML structure |
Family relationship tree
Performance Score Distribution (Top 20)
Click a model name to view its detail page.
Price-Performance Score Distribution (Top 20)
Click a model name to view its detail page.
| Score | Cost | Time | ||
|---|---|---|---|---|
| Gemini 4 Argon (Reasoning, Medium) | 93% | $0.056 | 26.0s | |
| Gemini 3.7 Flash (Reasoning, Medium) | 92% | $0.048 | 30.7s | |
| GPT-6.1 Sol (Reasoning, Medium) | 94% | $0.063 | 1.3m | |
| GPT-6 Sol (Reasoning, Medium) | 93% | $0.073 | 1.1m | |
| Gemini 3.8 Flash (Reasoning, Medium) | 94% | $0.087 | 53.4s | |
| DeepSeek V4.1 Flash 1017 (Reasoning, High) | 88% | $0.0082 | 1.3m | |
| GPT-6 Sol | 88% | $0.051 | 30.2s | |
| GPT-6 Luna (Reasoning, High) | 91% | $0.0082 | 2.1m | |
| GPT-6 Luna (Reasoning, Medium) | 88% | $0.0050 | 1.7m | |
| Claude Haiku 5.5 (Adaptive) | 84% | $0.0090 | 40.7s | |
| DeepSeek V4.1 Flash (Reasoning, High) | 89% | $0.019 | 2.0m | |
| GPT-5.6 Terra (Reasoning, Medium) | 86% | $0.082 | 32.1s | |
| Gemini 3.6 Flash (Reasoning) | 91% | $0.114 | 48.1s | |
| Gemini 2.5 Flash | 78% | $0.010 | 6.9s | |
| Claude Haiku 5.5 (Reasoning, Low) | 81% | $0.0067 | 24.7s | |
| Muse Spark 1.1 (Reasoning, Medium) | 85% | $0.055 | 43.3s | |
| Gemini 3.1 Flash Lite (Preview) | 78% | $0.0062 | 3.9s | |
| Muse Wonder (Reasoning, Medium) | 91% | $0.153 | 25.7s | |
| Grok 4.5 (Reasoning, Low) | 82% | $0.046 | 48.7s | |
| GPT-5.4 (Reasoning, Low) | 85% | $0.072 | 50.1s | |
Most Stable Models (Top 20)
Ranked by stability (median × consistency). Click a model name to view its detail page.
| Score | Consistency | Stability | ||
|---|---|---|---|---|
| GPT-6.1 Sol (Reasoning, Medium) | 94% | 100% | 93% | |
| Claude Opus 5.5 (Reasoning) | 93% | 99% | 93% | |
| Gemini 4 Argon (Reasoning, Medium) | 93% | 100% | 93% | |
| GPT-5.6 Sol (Reasoning, Medium) | 93% | 99% | 92% | |
| GPT-6 Sol (Reasoning, Medium) | 93% | 98% | 92% | |
| Gemini 3.8 Flash (Reasoning, Medium) | 94% | 98% | 91% | |
| Grok 4.7 (Reasoning, High) | 92% | 98% | 91% | |
| Claude Opus 4.6 (Reasoning) | 92% | 98% | 90% | |
| Gemini 3.7 Flash (Reasoning, Medium) | 92% | 97% | 90% | |
| Ember-1 (Reasoning, Max) | 94% | 96% | 90% | |
| Grok 4.6 (Reasoning, High) | 91% | 96% | 89% | |
| GPT-6 Luna (Reasoning, High) | 91% | 97% | 89% | |
| Gemini 3.6 Flash (Reasoning) | 91% | 97% | 88% | |
| GPT-5.4 (Reasoning) | 93% | 95% | 88% | |
| GPT-5.5 (Reasoning, Medium) | 90% | 98% | 88% | |
| Grok 4.5 (Reasoning, High) | 89% | 99% | 88% | |
| Muse Wonder (Reasoning, Medium) | 91% | 95% | 87% | |
| Xiaomi MIMO v2.6 Flash | 91% | 94% | 85% | |
| GPT-6 Luna (Reasoning, Medium) | 88% | 97% | 85% | |
| Claude Opus 4.6 | 88% | 97% | 85% | |
Top Overall Models (Top 20)
Ranked by composite score (performance, cost, speed & stability). Click a model name to view its detail page.
| Score | Cost | Speed | Stability | ||
|---|---|---|---|---|---|
| Gemini 4 Argon (Reasoning, Medium) | 93% | $0.056 | 26.0s | 93% | |
| Gemini 3.7 Flash (Reasoning, Medium) | 92% | $0.048 | 30.7s | 90% | |
| GPT-6.1 Sol (Reasoning, Medium) | 94% | $0.063 | 1.3m | 93% | |
| GPT-6 Sol (Reasoning, Medium) | 93% | $0.073 | 1.1m | 92% | |
| Gemini 3.8 Flash (Reasoning, Medium) | 94% | $0.087 | 53.4s | 91% | |
| GPT-6 Luna (Reasoning, High) | 91% | $0.0082 | 2.1m | 89% | |
| Gemini 3.6 Flash (Reasoning) | 91% | $0.114 | 48.1s | 88% | |
| GPT-6 Luna (Reasoning, Medium) | 88% | $0.0050 | 1.7m | 85% | |
| GPT-6 Sol | 88% | $0.051 | 30.2s | 83% | |
| Muse Wonder (Reasoning, Medium) | 91% | $0.153 | 25.7s | 87% | |
| DeepSeek V4.1 Flash 1017 (Reasoning, High) | 88% | $0.0082 | 1.3m | 80% | |
| DeepSeek V4.1 Flash (Reasoning, High) | 89% | $0.019 | 2.0m | 82% | |
| Muse Spark 1.1 (Reasoning, Medium) | 85% | $0.055 | 43.3s | 83% | |
| Claude Sonnet 5.5 (Reasoning) | 90% | $0.148 | 41.9s | 85% | |
| Claude Haiku 5.5 (Adaptive) | 84% | $0.0090 | 40.7s | 79% | |
| GPT-5.4 (Reasoning, Low) | 85% | $0.072 | 50.1s | 83% | |
| Claude Haiku 5.5 (Reasoning, Low) | 81% | $0.0067 | 24.7s | 79% | |
| GPT-5.6 Sol (Reasoning, Medium) | 93% | $0.262 | 1.4m | 92% | |
| GPT-5.6 Terra (Reasoning, Medium) | 86% | $0.082 | 32.1s | 79% | |
| Kimi K3 (Reasoning, Low) | 86% | $0.081 | 1.7m | 84% | |
| Median | Evaluator | Top 3 | Flop 3 |
|---|---|---|---|
| 84.3% | Alias accuracy | ||
| 100.0% | Character precision | ||
| 100.0% | Character recall | ||
| 100.0% | Isolated character handling | ||
| 100.0% | Red-herring resistance | ||
| 17.7% | Relationship category recall | ||
| 100.0% | Relationship endpoint integrity | ||
| 70.7% | Relationship precision | ||
| 7.7% | Relationship recall | ||
| 85.1% | Relationship type accuracy | ||
| 100.0% | XML structure |