NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2024-07-19-01J35TATZGB362QQY1WAA49JVC
Run Details
Model:
DeepSeek-V2 Chat
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
17.5s
Cost:
$0.000091
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
9.61%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
4.8041%
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
Molly tapped her pen against the blueprint of the community center. "So, we hit the party at 7 PM sharp. The clown's act will be in full swing, and no one will notice a couple of extra hands in the gift pile." Bill, chewing on a stale pretzel, squinted at the map. "You sure about this, Molls? We're talking about kids' birthday loot here. Not exactly the vault of the Federal Reserve." She rolled her eyes. "It's not just any party. It's the mayor's daughter's bash. Trust me, the presents will be worth our while." "Alright, alright. But what's our cover?" Molly grinned, her eyes twinkling. "Easy. We go as... entertainers. You be the magician. I'll be the... balloon animal artist." Bill snorted. "Balloon animal artist? You can't even tie your shoes without a diagram." "Hey, I watched a YouTube tutorial. I'm practically a professional." "And I'm Houdini," Bill muttered, but a smile crept across his face. "Alright, let's get those balloons inflated and those rabbits ready." Molly clapped her hands. "Perfect. Operation 'Happy Unbirthday' is a go."
id
"gen-JAlSXcYMCvrrTzMLebPFNpsUO0vG"
total_cost
0.00009142
upstream_id
"0a2478042857436e94b176b5dbdb8b33"
created_at
"2024-07-19T15:39:12.076503+00:00"
7/19/2024, 3:39:12 PM
model
"deepseek/deepseek-chat"
app_id
182717
streamed
false
cancelled
false
provider_name
"DeepSeek"
latency
17464
moderation_latency
(null)
generation_time
0
finish_reason
"stop"
tokens_prompt
68
tokens_completion
261
native_tokens_prompt
73
native_tokens_completion
290
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.00009142
Evaluation details
Result
Evaluator
Details
Meta Data
9.61%
Matches word count
n/a
neededClean
false
words
178
0.00%
Dialogue to Total Word Ratio
Ratio: 74.72%, Deviation: 64.72%
neededClean
false
wordsTotal
178
wordsDialogue
133
4.8041%