Grape benchmark

Suites / English

English

28 questions on 12 English Wikipedia articles (tea, octopus, Marie Curie, ...), 630 KB. Four are asked in Hindi, Spanish or French, and five are off-topic.

3 runs / 12 files, 0.6 MB / model haiku, thinking off / Vector RAG embedder BAAI/bge-small-en-v1.5. Raw results: 2026-09-30 16:20 2026-09-30 16:23 2026-09-30 16:26

Grape

23 / 23

Off-topic refused
5 / 5
Input tokens per question
1,217 (1,203 to 1,244)
LLM calls per question
1.12 (1.11 to 1.14)
Cost per question
$0.00148 ($0.00145 to $0.00152)
Time per question
3.8 s (3.7 s to 3.8 s)
Index time
47 ms
Vector RAG

19 / 23

Off-topic refused
5 / 5
Input tokens per question
1,307
LLM calls per question
1.00
Cost per question
$0.00155 ($0.00155 to $0.00156)
Time per question
3.0 s (3.0 s to 3.1 s)
Embedding time
1.6 min
Questions

right every run some runs never. Left half Grape, right half Vector RAG.

  1. 01
  2. 02
  3. 03
  4. 04
  5. 05
  6. 06
  7. 07
  8. 08
  9. 09
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28

j / k or arrow keys