NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM61CWHK1JPTRF8PY205S71
Run Details
Model:
DeepSeek V3.1
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
2.0s
Cost:
$0.000494
Provider:
SambaNova
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 slapped the newspaper down on the table. "It's the Van der Clue mansion. They hired a clown. We need a plan." "A clown?" Bill scoffed. "What’s the angle?" "The angle is a ten-foot-tall bouncy castle, a distracted staff, and a guest list of toddlers. Their security is a guy named Carl who mostly worries about bubble solution spills. The haul is in a designated 'Gift Room.' It's a velvet-roped library." Bill stared. "You want us to crash a five-year-old’s birthday party." "It’s a surgical strike. We go in as caterers. You’re the miserable balloon-animal guy." "I don’t do balloon animals." "You'll learn. I'll handle the extraction. We stuff the loot into a giant, empty gift bag, waltz right out the front door during the pinata chaos." "And the kids? What if one cries?" "You give them a poorly-made balloon sword and run. This isn't 'Ocean's Eleven,' it's a cakewalk. The hardest part will be not eating all the mini hot dogs." Bill sighed, rubbing his temples. "Fine. But if a single child asks me for a poodle, I'm out. I have a reputation." "Your reputation is that you once got outsmarted by a motion-sensor light. Now, practice your twists. You need to master the poodle."
created_at
"2026-02-16T21:34:56.587641+00:00"
2/16/2026, 9:34:56 PM
model
"deepseek/deepseek-chat-v3.1"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
204
moderation_latency
(null)
generation_time
1924
tokens_prompt
95
tokens_completion
319
native_tokens_prompt
67
native_tokens_completion
300
native_tokens_completion_images
(null)
native_tokens_reasoning
0
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.00049355
router
(null)
provider_responses
0
id
"d247759a-8bdf-4af6-b590-40f20b9cf4d3"
status
200
is_byok
false
latency
204
endpoint_id
"9abe71b4-ff68-4f47-bf15-6cf4d76d9a80"
provider_name
"SambaNova"
model_permaslug
"deepseek/deepseek-chat-v3.1"
api_type
"completions"
id
"gen-1771277693-7Qu9MmzV79suZzUHjgpg"
upstream_id
"d247759a-8bdf-4af6-b590-40f20b9cf4d3"
total_cost
0.00049355
cache_discount
(null)
upstream_inference_cost
0
provider_name
"SambaNova"
Evaluation details
Result
Evaluator
Details
Meta Data
68.1022%
Matches word count
n/a
neededClean
false
words
214
99.8214%
Dialogue to Total Word Ratio
Ratio: 92.06%, Deviation: 2.06%
neededClean
false
wordsTotal
214
wordsDialogue
197
83.9618%