NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-25-01KQ2DQ4GB75M0ZGDC1M1E8QRZ
Run Details
Model:
DeepSeek V4 Flash
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
6.0s
Cost:
$0.000111
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
0.08%
Matches word count
(details)
36.79%
Dialogue to Total Word Ratio
(details)
18.4364%
Overall
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
“Alright, Bill, run it by me one more time.” “We go in as clowns. I do balloon animals, you do face paint. We blend in.” “Blend in? With these floppy shoes? And whose idea was the red nose?” “It’s camouflage, Molly. No one suspects a clown.” “I suspect every clown, Bill. They’re terrifying. But fine. What about the presents?” “The piñata takes the hit first. Kids scream, parents rush. We swap the gift bags with decoys.” “Decoys filled with what? Rocks?” “Socks. Clean, folded socks. They’ll think it’s a late Christmas miracle.” “Parents will be thrilled. Kids will burn the house down. Then what?” “We exit through the bounce house. Slow. Casual. Waving.” “Waving. At the children we just robbed.” “It’s called misdirection, Molly. Give ‘em a smile, they forget the hurt.” “You call that misdirection? I call it a lawsuit. What if the birthday kid spots us?” “Then you do a magic trick. Make him disappear. Poof.” “I don’t do magic, Bill. I steal. There’s a difference.” “Fine. We pivot. If caught, we act like sad clowns. Tears work wonders.” “I’m not crying over a stolen present. I have standards.” “Just dab the eyes. Big, wet, theatrical. They’ll feel so bad they’ll give us cake.” “Cake. You’re planning to eat cake at a kids’ party we’re robbing.” “It’s in the union handbook. Page four. ‘Always secure the refreshments.’”
created_at
"2026-04-25T13:36:57.747Z"
4/25/2026, 1:36:57 PM
model
"deepseek/deepseek-v4-flash-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
550
moderation_latency
(null)
generation_time
5475
tokens_prompt
77
tokens_completion
411
native_tokens_prompt
67
native_tokens_completion
364
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)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.0001113
router
(null)
provider_responses
0
endpoint_id
"722e0746-2034-4d6e-a08f-e6271849ee9e"
id
"f1804d96-f474-480d-837d-c7ededcb56ac"
is_byok
false
latency
347
model_permaslug
"deepseek/deepseek-v4-flash-20260423"
provider_name
"DeepSeek"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777124217-WFG321BSaNUziO7zcmiG"
session_id
(null)
api_type
"completions"
id
"gen-1777124217-DKFtQLk7ymg4y8tYKQr9"
upstream_id
"f1804d96-f474-480d-837d-c7ededcb56ac"
total_cost
0.0001113
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepSeek"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.08%
Matches word count
n/a
neededClean
false
words
229
36.79%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
229
wordsDialogue
229
18.4364%