Narev
  • Twitter
  • Register

  • Evals
    • Models
    • Hardware

  • Catalog

> Hub > Evals > Models > …

loading run

> Hub > Evals > Models > Zverev–Sonego US Open Match Details

These items measure recall and interpretation of the reported five-set US Open match, including its score, duration, turning point, and finishing time.

Scorer: Factuality

loading chart
Score vs cost per task
#scoremodel
cost
price
ttfttotal tok
150%perplexity/sonar-pro-search$0.000223.23s68
236%gemini-3.7-flash
$0.00296
$0.75 / $3.75 per 1M
3.94s846
3failedqwen/qwen3.8-flash
—
$0.15 / $0.47 per 1M
—0

Prompts

  • At approximately what time did the match finish? A) Just after midnight B) Almost 2 a.m. C) Almost 3…
  • At which tournament was the reported match played? A) Australian Open B) French Open C) Wimbledon D)…
  • How long was the match reported to have lasted? A) 3 hours 52 minutes B) 4 hours 25 minutes C) 4 hou…
  • How many sets were played in total? A) Two B) Three C) Four D) Five Reply with the letter of the cor…
  • In which set was Zverev reportedly two points away from losing the match? A) Fourth set B) First set…
  • Which scoreline was reported for the match from Zverev’s perspective? A) 6-4, 3-6, 6-7(7), 7-5, 6-4 …
  • Which set was decided by a tiebreak? A) First set B) Second set C) Third set D) Fifth set Reply with…
  • Which sets did Zverev win according to the reported scoreline? A) First, second, and fifth B) First,…
  • Who was identified as Alexander Zverev’s opponent in the match? A) Lorenzo Musetti B) Lorenzo Sonego…
  • Who won the third set, based on the reported scoreline from Zverev’s perspective? A) Zverev, 7-5 B) …
Create your own benchmark →