NC Bench
Overview
About
Tests
Models
Model
Claude 3.5 Sonnet
Test
Dialogue tags
Scenario
Write 200 words with 90% dialogue
Run ID
2024-07-19-01J35T09KQXT9KR0XM700E2P4Z
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, are you sure about this? It's a kids' party, for crying out loud!" "Bill, keep your voice down. And yes, I'm sure. Think of all the presents we can snag." "But... it's just wrong. They're children!" "Oh, please. These are rich kids. They'll get more presents next week." "I don't know..." "Look, do you want to make some easy money or not?" "I guess so, but—" "No buts. Here's the plan: we sneak in through the back door during the magic show." "Magic show?" "Yeah, they've hired some wannabe Houdini. Anyway, while everyone's distracted, we grab the gifts and bail." "What if we get caught?" "We won't. Trust me, I've done this before." "You've robbed a children's party before?" "Well, no, but how hard can it be?" "I've got a bad feeling about this, Molly." "Relax, Bill. It'll be a piece of cake. Speaking of which, we might even score some of that too." "You're impossible." "I prefer the term 'opportunistic.' Now, are you in or out?" "Fine, I'm in. But if this goes south, I'm blaming you." "Fair enough. Let's go steal from some kids!"
id
"gen-WY2lFZVDfBldCV7v7axHEK8C5vst"
total_cost
0.005055
upstream_id
"msg_019wFeZ2SDMBc3a2SwrbtrwC"
created_at
"2024-07-19T15:33:21.433426+00:00"
Friday, July 19th, 2024 at 3:33:21 PM GMT+00:00
model
"anthropic/claude-3.5-sonnet:beta"
app_id
182717
streamed
false
cancelled
false
provider_name
"Anthropic"
latency
12323
moderation_latency
(null)
generation_time
0
finish_reason
"end_turn"
tokens_prompt
68
tokens_completion
278
native_tokens_prompt
75
native_tokens_completion
322
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.005055
Evaluation details
Result
Evaluator
Details
Meta Data
75.1556%
Matches word count
n/a
neededClean
false
words
187
36.7879%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
187
wordsDialogue
187
55.9718%