NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-20-01M5C6P35QE59FA8F9DM86QYTJ
Run Details
Model:
DeepSeek V4.1 Flash 1017 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 9s
Cost:
$0.005744
Provider:
InferenceNet
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
97.61%
Dialogue to Total Word Ratio
(details)
98.8036%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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 unfolded the birthday-party flyer across the van’s dashboard and drew a red circle around the gazebo. Bill adjusted his ski mask, which had ridden up until it looked like a nervous headband, and studied the bouncy castle wheezing across the street. A clown was already trapped in the inflatable entrance, waving one oversized shoe for help. “Remind me why we’re robbing a seven-year-old,” Bill said. “Banks have guards, cameras, and chained pens,” Molly said. “Timmy’s party has a shoeless clown and a grandmother asleep under cake.” “That clown saw me.” “That clown saw a man in a ski mask and assumed you were entertainment.” “I’m not entertainment. I’m a professional.” Bill tapped the clipboard and flipped to a floor plan drawn on the back of a pizza menu. “Entry points?” “Front door, side gate, and the inflatable castle has a safety flap in the back.” “The flap is for safety.” “So is not robbing children, but we’ve made choices.” Molly pulled out a crumpled diagram and tapped the gazebo. “Phase one: you distract the clown. Ask for a dog, a wolf, then a wolf in a vest.” “That could take twenty minutes.” “Exactly. Phase two: I slip through the side gate, avoid cupcakes, and bag every present.” “What about the grandmother?” “She’s snoring in three-quarter time. We’re fine.” Bill squinted at the flyer. “No gifts over ten dollars.” “That’s why we’re taking all of them. Volume, Bill. It’s a numbers game.” “It’s a toy dinosaur game.” “Some dinosaurs are educational.” “What if the kids see us?” “They won’t. Kids notice only sugar and someone saying not now.” “What if they cry?” “You do the thing.” “I’m not doing the thing.” Molly checked her watch and pulled two party hats from the glove compartment. “We go in four minutes. You take the clown, I take the presents, and we meet at the van.” “The bouncy castle?” “What about it?” “I’ve always wanted to go in one. Reconnaissance.” “Fine. Two bounces. Keep the mask on.” “Three bounces.” “Two bounces and you carry the dinosaur.” “Deal. But I’m keeping the dinosaur if it’s educational.” Bill folded the clipboard, tucked it into his jacket, and opened the van door with solemnity. Molly secured the party hats, checked the sack for holes, and reminded herself that volume was a numbers game, even if the numbers were mostly plastic dinosaurs and gift cards to a yogurt shop. Across the street, the clown escaped the inflatable entrance and fell into the punch table. The grandmother did not stir. The plan was not elegant, but it had balloons, panic, and a backup sack for dinosaurs. The children screamed with joy that sounds identical to panic, which Molly considered useful cover. Bill took one step toward the party, then paused to straighten his mask. It slipped over his eyes. Molly steered him toward the side gate, toward the clown, toward the tiny vest, and toward the van, the engine running and getaway playlist set to children’s music.
created_at
"2026-10-20T08:26:13.16Z"
10/20/2026, 8:26:13 AM
model
"deepseek/deepseek-v4.1-flash-20261017"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
249
moderation_latency
(null)
generation_time
68361
tokens_prompt
75
tokens_completion
11500
native_tokens_prompt
67
native_tokens_completion
11478
native_tokens_completion_images
(null)
native_tokens_reasoning
10777
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
0.00574436
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1792484773-wqaQeNtfAhX5sXK8yytF"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1792484773-rBKfmXy3KSZo968xbD2b"
upstream_id
"gv2_ZMWa24fg1YWoEVR5wIS2y"
provider_responses
0
endpoint_id
"4231aea9-73af-4a2f-bd14-660b1d8b13b2"
id
"gv2_ZMWa24fg1YWoEVR5wIS2y"
is_byok
false
latency
245
model_permaslug
"deepseek/deepseek-v4.1-flash-20261017"
provider_name
"InferenceNet"
status
200
total_cost
0.00574436
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
100.00%
Matches word count
n/a
neededClean
false
words
499
97.61%
Dialogue to Total Word Ratio
Ratio: 49.30%, Deviation: 0.70%
neededClean
false
wordsTotal
503
wordsDialogue
248
98.8036%