NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-20-01M5C6NTXY03Z44TZBEZQQ98KG
Run Details
Model:
DeepSeek V4.1 Flash 1017 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
43.5s
Cost:
$0.003965
Provider:
InferenceNet
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
97.96%
Dialogue to Total Word Ratio
(details)
48.9796%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 lowered the binoculars and stared at the party tent pitched behind the community center, where a banner announced HAPPY BIRTHDAY, TIMMY in letters that drooped in the heat. The van smelled of stale coffee and wet crayons. A plastic clown nose hung from the rearview mirror like a warning. "Walk me through it again, Bill. Last time you called a job child’s play, I ended up in a bouncy castle with a sprained ankle and a party hat glued to my head." Bill unfolded a crayon-covered napkin across the dashboard and smoothed it with the flat of his hand. It was covered in arrows, stars, and a crude drawing of a wagon. "We wait for the clown. Clowns are chaos. Nobody watches the presents when a man in size-twenty shoes is making a poodle out of balloons." "That poodle is not a permit," Molly said, drumming her fingers on the steering wheel. "What about the parents?" "Parents are easy. They’re all arguing about gluten by the cake table. We slip in as entertainment." "We are not entertainers." "We’re wearing magician suits." "Bill, you bought those at a costume shop that smelled like soup." "Exactly. Low expectations." Bill tapped the napkin. "Our first trick is the old vanishing gift table." Molly pinched the bridge of her nose and looked at the cluster of folding tables under the tent. The tent flaps breathed in the hot wind. "There are forty-seven presents, three gift bags, and one bicycle." "The bicycle complicates things." "You think?" "We use the wagon. Kids love wagons. We load it, wave at a toddler, and walk out like cleanup." "Cleanup crews don’t wear capes." "Fine. We lose the capes." "And the masks?" "The masks stay. It’s a superhero party." Molly groaned. "Do you know what happens when you steal from a superhero party? Six-year-olds chase you in capes. They have no fear and terrible aim with juice boxes." "Then we split up. You take the gift bags. I take the bicycle." "The bicycle has streamers." "Streamers are not a security system." "They’re a visual alert, Bill. Everyone sees a grown man riding a tiny pink bicycle with tassels." "Not if I ride it confidently." Molly checked the tent flap, where a cluster of parents had begun wrestling a piñata into position. "Why are we stealing from a children’s party? We used to rob casino vaults." "We used to rob casino vaults. Now we’re freelance. The client wants the presents." "Which client?" "The one who paid cash and said, ‘Don’t overthink it.’" "That’s not a client. That’s a red flag with a wallet." Bill tapped the napkin again and drew a line under the word CAKE. "Party starts at two. Clown at two-fifteen. Cake at two-thirty. Presents at three. We need to be in the van by three-oh-seven." "Why three-oh-seven?" "Because the magician’s finale is a smoke bomb. I bought one." "You bought a smoke bomb for a children’s party?" "It’s for us. The kids get a fake one. They’ll never know the difference." Molly sighed and rubbed her temples. "Fine. But if a child cries, I’m leaving." "Agreed." "And if a clown sees us?" Bill smiled. "Then we offer him a cut."
created_at
"2026-10-20T08:26:04.612Z"
10/20/2026, 8:26:04 AM
model
"deepseek/deepseek-v4.1-flash-20261017"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
108
moderation_latency
(null)
generation_time
43441
tokens_prompt
75
tokens_completion
7327
native_tokens_prompt
67
native_tokens_completion
7919
native_tokens_completion_images
(null)
native_tokens_reasoning
7168
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"
service_tier
(null)
usage
0.00396486
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1792484764-dsgTW9WcsAyy1tbSjQdP"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1792484764-UKOYJPU8d6qOytEzKOcf"
upstream_id
"gv2_eUZGozhhWKibRDhvdiJf0"
provider_responses
0
endpoint_id
"4231aea9-73af-4a2f-bd14-660b1d8b13b2"
id
"gv2_eUZGozhhWKibRDhvdiJf0"
is_byok
false
latency
108
model_permaslug
"deepseek/deepseek-v4.1-flash-20261017"
provider_name
"InferenceNet"
status
200
total_cost
0.00396486
cache_discount
(null)
upstream_inference_cost
0
provider_name
"InferenceNet"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
534
97.96%
Dialogue to Total Word Ratio
Ratio: 70.64%, Deviation: 0.64%
neededClean
false
wordsTotal
545
wordsDialogue
385
48.9796%