NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-20-01M5C6BGK354GTCYFEP4J5T1JR
Run Details
Model:
DeepSeek V4.1 Flash 1017 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
42.5s
Cost:
$0.003342
Provider:
InferenceNet
Evaluation summary
Score
Evaluator
Details Link
86.38%
Matches word count
(details)
84.34%
Dialogue to Total Word Ratio
(details)
85.3584%
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
Bill flattened a crayon-drawn map of the Petting Zoo Pavilion across the picnic table, weighting the corners with a juice box, a plastic dinosaur, and a half-eaten granola bar. Rain drummed on the umbrella. Somewhere behind the hedge, a party horn honked mournfully like a goose being stepped on. A toddler in a ladybug costume pressed her face against the umbrella, stared at the map, and then screamed with the pure, unfocused fury of someone who had already eaten too much frosting. Bill slid a rubber snake under the granola bar and tapped the gazebo tent. "Party starts at two. We go in as entertainment." Molly squinted at the map. "We are not clowns." "We're magicians. I brought a cape." "That cape smells like a hamster. No." "Fine. Face painters. We wear smocks, blend in, and when the piƱata explodes, we scoop the presents into the diaper bag." Molly pulled a pencil from behind her ear and circled the bouncy castle. "The diaper bag is already full of diapers." "Decoy diapers. Very professional." "They're real diapers, Bill." "Real diapers are the best decoy. Nobody searches a diaper bag. It's the law." "Nobody searches it because it's disgusting. What's the entry?" Bill grinned and drew a dotted line through the balloon arch. "Seven-foot clearance, one security guard named Gary, and a bouncy castle blocking the east exit. Gary will recognize me." "So don't wear the cape." "The cape is non-negotiable. It has pockets." Molly wrote NO JUICE BOX in large letters. "Presents are behind the unicorn throne. We need to get past a princess, a pirate, and eleven toddlers on sugar." "Toddlers are the real threat. They have no financial motivation and no fear of prison." "Exactly. So we use the snake." Bill froze. "We do not have a snake." "The balloon snake. You inflate it during the animal parade. I tell everyone it's magic. Chaos. We grab the presents, load them into the stroller, and walk out like responsible parents." "Whose stroller?" "The one with the sleeping baby." "Absolutely not." "We put the baby in the bouncy castle first." "That's worse!" "Bill, focus. We are borrowing the stroller from the baby. There's a difference." "There is literally no difference." Molly sighed. "Fine. We bring our own stroller." "With a fake baby?" "With a fake baby and a real diaper bag full of real diapers, because apparently that's our brand now." Bill nodded slowly. "I like it. What about Gary?" "I distract Gary with a lost child." "We don't have a lost child." "We have a magician's rabbit and a cape. The rabbit becomes the lost child." "That doesn't make sense." "Neither does wearing a cape to a toddler's birthday, but here we are." Bill folded the map and tucked it into the cape's suspiciously deep pocket. "Okay. Two o'clock. Smocks, balloon snake, fake baby, decoy diapers. No clowns." "No clowns," Molly said. "And Bill?" "Yes?" "Bring the juice box."
created_at
"2026-10-20T08:20:26.345Z"
10/20/2026, 8:20:26 AM
model
"deepseek/deepseek-v4.1-flash-20261017"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
328
moderation_latency
(null)
generation_time
42441
tokens_prompt
75
tokens_completion
6208
native_tokens_prompt
67
native_tokens_completion
6673
native_tokens_completion_images
(null)
native_tokens_reasoning
5979
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
0.00334186
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1792484426-mg4RXFrbo883lvtancT8"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1792484426-nySA3lLXozwIrTARX6ic"
upstream_id
"gv2_WgeXkIIgwqfIFvTjarqpk"
provider_responses
0
endpoint_id
"4231aea9-73af-4a2f-bd14-660b1d8b13b2"
id
"gv2_WgeXkIIgwqfIFvTjarqpk"
is_byok
false
latency
328
model_permaslug
"deepseek/deepseek-v4.1-flash-20261017"
provider_name
"InferenceNet"
status
200
total_cost
0.00334186
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
86.38%
Matches word count
n/a
neededClean
false
words
489
84.34%
Dialogue to Total Word Ratio
Ratio: 68.15%, Deviation: 1.85%
neededClean
false
wordsTotal
493
wordsDialogue
336
85.3584%