NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-20-01M5C6K5QXT1XEGE0B82SS43DN
Run Details
Model:
DeepSeek V4.1 Flash 1017 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
16.5s
Cost:
$0.000978
Provider:
InferenceNet
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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 spread the birthday invitation on the van’s dashboard like a mafia map. “Right, Bill. The target is Little Timmy’s seventh birthday party, Saturday, two o’clock, community hall, balloons, cake, and a present table with an estimated retail value of four thousand pounds.” Bill squinted. “Four thousand? For a seven-year-old?” “His dad’s a dentist and his mum’s got guilt issues. There’s a scooter, a drone, three Lego sets, and a gift card situation.” “I don’t like it. Kids are unpredictable. One of them cried at the last job and I still hear it.” “That was you, Bill. You cried because the guard’s dog looked sad.” “He did look sad. Anyway, what’s the cover?” “We go in as entertainers. You’re Mr. Fizzbang, a children’s magician. I’m your assistant, Sparkle.” “I don’t do magic.” “You do one trick: you make presents disappear.” “That’s theft.” “It’s misdirection. While you distract them with a fake rabbit, I’ll bag the loot.” “Where do we get a fake rabbit?” “I’ve got a real one.” “Molly, no.” “He’s very calm. His name’s Vincent. He owes me a favour.” “A rabbit owes you a favour?” “He ate my parking ticket. We’re square.” Bill rubbed his temples. “What about the parents?” “Two mums on prosecco, one dad filming everything, and a grandmother who will absolutely notice a scooter leaving.” “So we need a diversion.” “Cake fight.” “We can’t start a cake fight at a child’s birthday.” “We don’t start it. We encourage it. You do the trick where you accidentally launch the cake.” “I knew my lack of magic would become a problem.” “Bill, focus. We need three minutes. You gather them for the show. I swap the presents into the laundry cart. Vincent does a hop. We exit as clowns.” “Why clowns?” “Because nobody questions clowns carrying a laundry cart.” “They question everything about clowns.” “Fine, elves. Christmas elves delivering a surprise.” “In July?” “It’s a themed party.” “Is it?” “It is now. I printed new invitations.” Bill stared. “You changed the invitations?” “Only the theme. The time and place are identical.” “And the parents won’t mind?” “They’ll be delighted. Everyone loves elves.” “What if the kids want to keep the presents?” “They will. That’s why we take them before cake.” “Before cake? That’s monstrous.” “After cake they’re sticky and suspicious. Before cake they’re innocent and gullible.” Bill checked a clipboard. “What about security?” “A bouncy castle attendant named Kevin. He’s seventeen and terrified of balloons.” “Aren’t we all.” “You are. That’s why you’re Mr. Fizzbang.” “And the getaway?” “Vincent’s cage in the back. We leave under a tablecloth. You honk.” “I honk?” “Twice. Once for danger, once for triumph.” “How do I tell them apart?” “You’ll know. Triumph feels expensive.”
created_at
"2026-10-20T08:24:37.378Z"
10/20/2026, 8:24:37 AM
model
"deepseek/deepseek-v4.1-flash-20261017"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
317
moderation_latency
(null)
generation_time
16516
tokens_prompt
75
tokens_completion
2180
native_tokens_prompt
67
native_tokens_completion
1945
native_tokens_completion_images
(null)
native_tokens_reasoning
1220
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
0.00097786
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1792484677-EpkKBKLBZj8Gn3iwTLED"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1792484677-acrRFq24VrfE9sUQrQw9"
upstream_id
"gv2_ojAlg6V3kCU1lGRtKIab4"
provider_responses
0
endpoint_id
"4231aea9-73af-4a2f-bd14-660b1d8b13b2"
id
"gv2_ojAlg6V3kCU1lGRtKIab4"
is_byok
false
latency
316
model_permaslug
"deepseek/deepseek-v4.1-flash-20261017"
provider_name
"InferenceNet"
status
200
total_cost
0.00097786
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
453
0.00%
Dialogue to Total Word Ratio
Ratio: 94.51%, Deviation: 24.51%
neededClean
false
wordsTotal
455
wordsDialogue
430
0.0000%