NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KC4DB94TT7DE75PW5D8P2P
Run Details
Model:
DeepSeek V4 Pro 0813 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
29.0s
Cost:
$0.004943
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
0.21%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.1071%
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 squinted at the inflatable bouncy castle. "Okay, run it again. We go in as what?" Bill adjusted his clip-on bow tie. "Party clowns. Obviously. You're Bubbles, I'm Honk." "Honk? That's not a clown name, that's a goose with a head injury." "It's whimsical. Kids love geese." "They don't. But fine. We walk in, we do balloon animals, then what?" "Then I create a diversion. I announce the magician has arrived. That's me, by the way. I do one trick. Kids are mesmerized." "You can't do one trick, Bill. Last time you tried to pull a quarter from behind a toddler's ear, you dropped a crowbar." "That's part of the charm. While they're all watching me, you start loading the gift table into the laundry cart." Molly rubbed her temples. "The laundry cart you stole from the nursing home?" "It has wheels, Molly. Wheels are wheels." "It smells like menthol and regret. And it's bright blue with 'Property of Shady Pines' stenciled on the side." "I covered it with a tarp. It says 'Party Supplies.' I used a sharpie. Very convincing." Molly stared at him. "You wrote 'Party Supplies' on a tarp. On top of a stolen nursing home cart. At a children's birthday party." "Exactly. No one questions a man with a tarp." "They question a man with a tarp who is sweating through a rainbow wig." Bill shrugged. "The wig is hot. Sue me." "What about the parents?" "Parents are exhausted. They'll be mainlining sheet cake and ignoring everything. Trust me. I've been to a birthday party." "When?" "Once. I think. There was a petting zoo." Molly sighed. "Fine. We're in. We've got the cart. How do we get out?" "Through the side gate. I'll tell them I need to retrieve my emotional support ferret from the van. Then I'll open the gate from the outside." "You don't have a ferret." "I have a sock with googly eyes. It's very believable." "Bill, this is the worst plan you've ever had." "Worse than the time we stole a Zamboni?" "That was you. I was the lookout." "Exactly. And you looked out beautifully." Molly shook her head. "What if a kid sees us loading the presents?" "Kids only care about sugar and violence. I'll hand them each a kazoo. Chaos will do the rest." "A kazoo? You want to arm children with kazoos?" "It's a diversion within a diversion. Like a heist onion." Molly looked at the gift table piled high with brightly wrapped boxes. "That's a lot of presents. What do we even do with a child's gifts?" "Return them for cash. Gift receipts, Molly. This is a professional operation." "Professional clowns named Bubbles and Honk, stealing from a six-year-old." "Seven. And it's not stealing. It's reallocating birthday joy to those more deserving. Us." Molly adjusted her rainbow suspenders. "I hate that I'm doing this." Bill grinned. "That's the spirit. Now honk your nose and let's go." "I'm not honking my nose." "You have to. It's part of the cover." Molly squeezed the red ball on her nose. It honked weakly. Somewhere in the backyard, a child screamed with joy. Bill smiled. "See? We're already blending in."
created_at
"2026-09-03T10:13:18.837Z"
9/3/2026, 10:13:18 AM
model
"deepseek/deepseek-v4-pro-20260813"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
226
moderation_latency
(null)
generation_time
28955
tokens_prompt
75
tokens_completion
1876
native_tokens_prompt
146
native_tokens_completion
1828
native_tokens_completion_images
(null)
native_tokens_reasoning
1057
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
0.0049426
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788430398-z3tVKiw4l67nlia99xre"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788430398-O3mZvfm5KxmHAsXupEbr"
upstream_id
"chatcmpl-RWwM6WZCC2IcutQ557XXV4Pl"
provider_responses
0
endpoint_id
"a69687f5-7209-4d1d-b4d6-50a9939aa5ce"
id
"chatcmpl-RWwM6WZCC2IcutQ557XXV4Pl"
is_byok
false
latency
144
model_permaslug
"deepseek/deepseek-v4-pro-20260813"
provider_name
"DeepInfra"
status
200
total_cost
0.0049426
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.21%
Matches word count
n/a
neededClean
false
words
528
0.00%
Dialogue to Total Word Ratio
Ratio: 86.63%, Deviation: 16.63%
neededClean
false
wordsTotal
531
wordsDialogue
460
0.1071%