NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1M01BNZMPXBJ1VVRG42XS35
Run Details
Model:
Hy4 Preview (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
9m 50s
Cost:
$0.116026
Provider:
Tencent
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
99.9990%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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
The van smelled like old fries and regret. Molly crouched behind the sliding door, binoculars pressed to her face. Across the street, a castle glowed with fairy lights. A banner above the porch read Happy Birthday, Timmy! Under it sat a mountain of presents. Inside, six-year-olds screamed like children hopped on cake. Molly’s eyes did the math before her conscience objected. That pile meant four months of rent, or one escape to a country without extradition. “This is a bad idea,” Bill said, crumbs falling from his chin. He sat on the van floor surrounded by empty chip bags and a duffel bag labeled Do Not Open Unless You Enjoy Consequences. Molly ignored him. Bill had called every job a bad idea since the petting zoo heist, which mostly involved a llama, a stroller, and a very loud goose. Molly lowered the binoculars. “Bad ideas pay the electric bill,” she said. “Look at that bike. That’s a mortgage payment with training wheels. Besides, Timmy’s mum bought out the entire toy aisle. It’s basically self-defense.” “We can’t steal from a birthday party,” Bill said. “It’s basically a felony with balloons.” He peered through the crack in the door as if hoping the presents would vanish and spare him the moral dilemma. Molly gave him the patient smile she reserved for bank managers and her cousin Derek. “We’re liberating presents from excess,” she said. “Timmy has two parties. He won’t notice. While you distract everyone, I slip through the kitchen, load the van, and we leave before the cake.” Bill opened his mouth, closed it, and opened it again. It was the same expression he had worn during the ice cream truck job, right before they escaped with forty dollars in quarters and a broken speaker. “How are we getting in?” he asked. “Through the front door? While a clown watches?” Molly reached into the duffel bag and pulled out a sequined cape, a plastic wand, and a hat that had seen better decades. “You’re the magician,” she announced. “The real one cancelled. I called as his assistant.” He would arrive, do a trick, and create chaos. Bill stared at the cape as if it had insulted him. “I don’t know any magic,” he said. “Last time I pulled a rabbit from a hat, I got a raccoon with very strong opinions about personal space.” Molly stuffed the cape into his hands. “Do the stapler trick,” she said. “Glitter covers everything.” While they screamed for cake, Molly would load the van and they would vanish like a bad investment. Bill’s sigh could have powered a small wind turbine. He draped the cape over his shoulders anyway, because he always did. “Fine,” he said. “But if a clown bites me, I’m billing you.” If he ended up covered in frosting, he would blame Molly in his memoir. Molly checked her watch, then the porch, then Bill’s cape for glitter, because priorities. “Noted,” she said, opening the van door. “Let’s go rob a childhood.”
created_at
"2026-09-03T16:01:10.346Z"
9/3/2026, 4:01:10 PM
model
"tencent/hy4-preview-20260827"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1609
moderation_latency
(null)
generation_time
589835
tokens_prompt
75
tokens_completion
38882
native_tokens_prompt
83
native_tokens_completion
46364
native_tokens_completion_images
(null)
native_tokens_reasoning
45670
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
0.116025586
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788451270-ClMju7PBji1qmPy2sD8V"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788451270-jES5wLClIDWx4ogti6p9"
upstream_id
"35f4463c-44cf-406e-942a-2eecfa794f57"
provider_responses
0
endpoint_id
"ad2a4093-3d2a-41d9-adf1-8b499ac8445d"
id
"35f4463c-44cf-406e-942a-2eecfa794f57"
is_byok
false
latency
1609
model_permaslug
"tencent/hy4-preview-20260827"
provider_name
"Tencent"
status
200
total_cost
0.116025586
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Tencent"
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
500
100.00%
Dialogue to Total Word Ratio
Ratio: 30.02%, Deviation: 0.02%
neededClean
false
wordsTotal
503
wordsDialogue
151
99.9990%