NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36SGDAR817YAJX6ZXB0FMRY
Run Details
Model:
Xiaomi MIMO v2.6 Pro
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
29.6s
Cost:
$0.000645
Provider:
Xiaomi
Evaluation summary
Score
Evaluator
Details Link
60.28%
Matches word count
(details)
0.53%
Dialogue to Total Word Ratio
(details)
30.4045%
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
Molly unfolded the napkin on the table between them and smoothed it flat with the edge of her hand. The napkin was covered in a hand-drawn floor plan of the Peterson residence, complete with arrows, stars, and what appeared to be a doodle of a cat wearing sunglasses. She tapped a star near the dining room. "That's where the presents pile up," she said. "Right beside the clown balloon arch." "Killer's name is Dexter. He goes by Dex." Bill leaned forward and squinted at the napkin as though it were an ancient map to buried treasure. He adjusted the tiny reading glasses perched on his nose, which made him look less like a career criminal and more like a confused librarian. "We're stealing from children." "We're stealing from parents who went overboard. The Johnsons alone spent four hundred on that Power Ranger sword." "Still feels wrong." Molly slid a second napkin across the table. This one featured a timeline written in green marker, annotated with little hearts and the word "GO" circled three times. She began walking him through the plan with the calm authority of a general briefing a reluctant soldier. The party was scheduled for two o'clock on Saturday, hosted at the Peterson home on Elm Street. Thirty-one children were expected. The children would be herded into the backyard for games around two-thirty, leaving the living room and dining room unmanned for at least twenty minutes while the parents supervised. Molly had confirmed this through a week of surveillance disguised as dog-walking and one memorable afternoon pretending to be a lost ice cream truck driver. Bill took a bite of his sandwich and chewed slowly while Molly described their disguises. He would pose as a balloon artist, because his hands were too large and clumsy to pass as a caterer, and Molly would don the uniform of a professional party clown. The beauty of the scheme lay in its simplicity. Nobody questioned a clown at a birthday party. Nobody ever did. It was the perfect camouflage, the kind of criminal brilliance that came naturally to Molly and made Bill deeply uncomfortable. "How much are we talking here?" Bill asked. "Roughly three grand in gifts, if my estimates are right. We unload the Power Ranger sword to Gary by Monday, and the tablet goes to that fence in Montrose." Bill swallowed his bite of sandwich. "And if Dex catches us?" "Dex is a balloon artist, not a cop. We tie him up with his own balloon animals." Bill stared at her for a long moment and then, despite every instinct telling him this was insane, began laughing. Molly folded the napkins and slipped them into her jacket pocket with the satisfaction of a woman who had just solved every problem in the world. The Petersen residence had no alarm system, one locked window on the east side, and a dog that Molly had already befriended with turkey slices during reconnaissance. Everything was in place. All that remained was for Saturday to arrive and for Bill to stop being a baby about stealing from six-year-olds.
created_at
"2026-09-23T09:28:19.553Z"
9/23/2026, 9:28:19 AM
model
"xiaomi/mimo-v2.6-pro-20260921"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
722
moderation_latency
(null)
generation_time
29526
tokens_prompt
75
tokens_completion
838
native_tokens_prompt
71
native_tokens_completion
706
native_tokens_completion_images
(null)
native_tokens_reasoning
84
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
0.000645105
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790155699-01elfdWTURIyP5Uf7mLe"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790155699-9N5asoTGLRDuOl0PEdcf"
upstream_id
"e61c0639-77c1-49f7-954a-8a389a2d78f1_80a3acb7846e4a31b3bffc060a084106"
provider_responses
0
endpoint_id
"4354fc36-b8ab-4ced-aa8c-0efb6b4523f1"
id
"e61c0639-77c1-49f7-954a-8a389a2d78f1_80a3acb7846e4a31b3bffc060a084106"
is_byok
false
latency
642
model_permaslug
"xiaomi/mimo-v2.6-pro-20260921"
provider_name
"Xiaomi"
status
200
total_cost
0.000645105
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Xiaomi"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
60.28%
Matches word count
n/a
neededClean
false
words
515
0.53%
Dialogue to Total Word Ratio
Ratio: 19.77%, Deviation: 10.23%
neededClean
false
wordsTotal
521
wordsDialogue
103
30.4045%