| # | Backend | Discovery Model | Extraction Model | KYS ↓ | Entities | Relations | Graph Size | Knowledge Rate | Total Time | Quality Score | Speed Score |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | anthropic | claude-sonnet-4-6 | claude-haiku-4-5-20251001 | 87 | 73 | 160 | 1.66 /s | 96.5s | 0.919 | 0.888 | |
| 2 | openai | gpt-4o | gpt-4o-mini | 47 | 52 | 99 | 1.16 /s | 85.7s | 0.569 | 1.000 | |
| 3 | ollama | llama3.2:latest | llama3.2:latest | 88 | 86 | 174 | 0.68 /s | 254.1s | 1.000 | 0.337 | |
| 4 | ollama | gemma3:12b | gemma3:4b | 94 | 34 | 128 | 0.68 /s | 187.2s | 0.736 | 0.457 | |
| 5 | ollama | llama3.1:8b | llama3.1:8b | 93 | 80 | 173 | 0.42 /s | 410.2s | 0.994 | 0.209 | |
| 6 | ollama | mistral:latest | mistral:latest | 46 | 35 | 81 | 0.16 /s | 510.6s | 0.466 | 0.168 | |
| 7 | ollama | qwen2.5:14b | qwen2.5:3b | 27 | 28 | 55 | 0.14 /s | 398.0s | 0.316 | 0.215 |
The Knowledge Yield Score (KYS) is a composite metric that balances output richness (how much the pipeline extracted) against efficiency (how fast it ran). It is the geometric mean of two normalized sub-scores, analogous to the F₁ score — it penalises runs that excel in only one dimension.
quality_norm rewards pipelines that produce large,
dense knowledge graphs. speed_norm rewards pipelines
that finish quickly; the fastest run scores 1.0 and slower runs are
penalised proportionally. The geometric mean ensures a pipeline
cannot compensate for poor speed with high quality alone — both
dimensions must be strong for a high KYS.
Knowledge Rate (graph_size / total_time) is a complementary raw throughput metric expressed in graph
elements per second, useful for absolute comparisons independent of
the normalization range.