NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1K9VM42X4MS4Z7EATZE9GQF
Run Details
Model:
DeepSeek V4 Flash 0731 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
6.0s
Cost:
$0.000114
Provider:
Wafer
Evaluation summary
Score
Evaluator
Details Link
86.38%
Matches word count
(details)
82.81%
Dialogue to Total Word Ratio
(details)
84.5967%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
“You sure about this, Bill?” Molly squinted at the rainbow banner strung across the backyard. “It’s a five-year-old’s party. The worst security is a piñata.” “That’s exactly why it’s brilliant,” Bill whispered, adjusting his fake mustache. “No one expects a heist from a clown.” “We’re not clowns. We’re wearing hoodies and tool belts. You look like a plumbing emergency.” “Plumbers attend parties. Kids love pipes.” Molly sighed. “Remind me the plan.” “I distract the parents with a balloon-animal giraffe. You slip into the gift tent, empty every box of Lego, switch the Barbies for rocks.” “Rocks? Where do I get rocks?” “I pre-loaded them in a diaper bag.” He tapped his bulge. “Fake poop on top. Zero inspection.” “Genius. And the cake?” “We’re stealing presents, not cake. That’s morbid.” “We’re stealing from a four-year-old!” “We’re liberating commercial excess. Philosophically.” “Philosophically, Bill, you just tripped over the bouncy castle.” “No—that’s part of the act. Looks like clumsiness.” He winced as a toddler grasp his ankle. “Go now. He’s crying. That’s my cue.” Molly took a breath. “Fine, but if I get face-painted, you’re handling the smash cake.” “Deal. But if you see a unicorn balloon, grab it. It’s probably worth ten Lego sets.” “You're insufferable.” She smiled. “Let’s take a birthday’s worth.”
created_at
"2026-09-03T09:33:33.712Z"
9/3/2026, 9:33:33 AM
model
"deepseek/deepseek-v4-flash-20260731"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2256
moderation_latency
(null)
generation_time
5943
tokens_prompt
75
tokens_completion
453
native_tokens_prompt
89
native_tokens_completion
422
native_tokens_completion_images
(null)
native_tokens_reasoning
59
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
0.0001144
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788428013-9Wh4kLVOjuz0oy5aAg6v"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788428013-Ve0EQdpAgAQStulZvMhG"
upstream_id
"3fe42821aa5241caada3e31ad320b961"
provider_responses
0
endpoint_id
"4d5286fc-8f37-40c2-a518-e1ae100d45fc"
id
"3fe42821aa5241caada3e31ad320b961"
is_byok
false
latency
2252
model_permaslug
"deepseek/deepseek-v4-flash-20260731"
provider_name
"Wafer"
status
200
total_cost
0.0001144
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Wafer"
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
211
82.81%
Dialogue to Total Word Ratio
Ratio: 83.41%, Deviation: 6.59%
neededClean
false
wordsTotal
217
wordsDialogue
181
84.5967%