NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2024-08-06-01J4MXXPNXA5H89A74JDSKY7Y4
Run Details
Model:
GPT-4o, Aug. 6th (temp=1)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
3.6s
Cost:
$0.002907
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
99.74%
Matches word count
(details)
8.07%
Dialogue to Total Word Ratio
(details)
53.9056%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% words of dialogue between two thieves named Molly and Bill who are planning to break into a children's party to steal all the presents. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
The sun was setting as Molly and Bill crouched behind the giant inflatable castle, observing the chaos of a sugar-fueled children's birthday party. Balloons bobbed gently in the breeze, a vivid gateway to their imminent haul. Molly squinted through her binoculars, calculating the odds amidst the giggling hordes. "You sure you can fit inside that panda costume, Bill?" she whispered, barely concealing her amusement. "Hey, it’s not the costume, it’s my charisma that’ll get us in," Bill replied, tugging at his panda head and doing a mock jig. He could almost hear the children's applause for his soon-to-be stellar performance. A burst of laughter erupted as another wave of hyper kids charged through the field, oblivious to the plotting pair. Molly tapped a finger against her chin. "Remember, focus on the big gifts. The little ones are decoys." "Right. Big boxes, big prizes," Bill nodded, adjusting his fur-covered sleeves as he mentally rehearsed his moves. The stakes were high. Not jewels, nor money, but an adrenaline rush wrapped in brightly colored paper. As the cake was being served, Molly nudged Bill. "Time to move, Panda Man." Bill winked and, with a puff of his fake paw, slipped toward the entrance. "Let’s unwrap some chaos."
id
"gen-BRrfy7MXWSgf2MH69vfDwuTFrsoa"
total_cost
0.0029075
upstream_id
"chatcmpl-9tNEDWcsF5fvrNqsVWwoBkBeF1jU6"
created_at
"2024-08-06T22:46:00.240409+00:00"
8/6/2024, 10:46:00 PM
model
"openai/gpt-4o-2024-08-06"
app_id
182717
streamed
false
cancelled
false
provider_name
"OpenAI"
latency
3264
moderation_latency
189
generation_time
0
finish_reason
"stop"
tokens_prompt
68
tokens_completion
279
native_tokens_prompt
67
native_tokens_completion
274
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.0029075
Evaluation details
Result
Evaluator
Details
Meta Data
99.74%
Matches word count
n/a
neededClean
false
words
204
8.07%
Dialogue to Total Word Ratio
Ratio: 22.60%, Deviation: 12.60%
neededClean
false
wordsTotal
208
wordsDialogue
47
53.9056%