NC Bench
Overview
About
Tests
Models
Model
DeepSeek-V2 Chat
Test
Dialogue tags
Scenario
Write 200 words with 90% dialogue
Run ID
2024-07-19-01J35SP5BP992M9KX8T1H8S8CG
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
Molly: "So, Bill, we're hitting the kiddie bash, right? Steal all those shiny presents?" Bill: "Yeah, Molly, but think about it. Kids' party, sugar and spice, and everything... not so nice." Molly: "Oh, come on, Bill. It's a goldmine! Toys, gadgets, maybe even a diamond tiara or two." Bill: "But Molly, we're talking about tots, tiny terrors, and their ticked-off parents." Molly: "Bill, we're pros. We blend in, grab the goods, and ghost before the first tantrum." Bill: "Blend in? How? We're not exactly Barney and Friends." Molly: "Easy, we dress up. You be the clown, I'll be the fairy godmother." Bill: "A clown? I look more like a serial killer in makeup." Molly: "It's perfect! They'll be too scared to notice the presents vanishing." Bill: "And you, a fairy? You can't even fly a kite." Molly: "Details, Bill. I'll dazzle 'em with glitter, you distract with juggling." Bill: "Juggling? I can barely juggle my own life." Molly: "Practice on oranges. We've got till Saturday." Bill: "Saturday? That's when the chaos hits?" Molly: "Exactly. Let the good times roll, and the presents... stroll." Bill: "Alright, Molly. But if this goes south, I'm blaming the fairy." Molly: "Deal. And if it goes smooth, I'll let you keep the tiara." Bill: "A tiara? Now you're speaking my language."
id
"gen-DL47tbL0xi4f1nfkC91JOX2M3bkU"
total_cost
0.00012446
upstream_id
"f588297c6db198d513383d69e9218d5b"
created_at
"2024-07-19T15:28:01.133077+00:00"
Friday, July 19th, 2024 at 3:28:01 PM GMT+00:00
model
"deepseek/deepseek-chat"
app_id
182717
streamed
false
cancelled
false
provider_name
"DeepSeek"
latency
24027
moderation_latency
(null)
generation_time
0
finish_reason
"stop"
tokens_prompt
68
tokens_completion
360
native_tokens_prompt
73
native_tokens_completion
408
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.00012446
Evaluation details
Result
Evaluator
Details
Meta Data
43.3783%
Matches word count
n/a
neededClean
false
words
217
99.9155%
Dialogue to Total Word Ratio
Ratio: 91.71%, Deviation: 1.71%
neededClean
false
wordsTotal
217
wordsDialogue
199
71.6469%