NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-07-01M4ASW9YWM5XHCQPHA51Y35GG
Run Details
Model:
Mistral Large 4.0 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
2m 53s
Cost:
$0.032282
Provider:
Mistral
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
17.72%
Dialogue to Total Word Ratio
(details)
8.8586%
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 fluorescent light in the back of the van buzzed like an angry bee, casting everything in a sickly yellow glow that made Bill's tactical vest look less like military surplus and more like a highlighter that had escaped a kindergarten. Molly sat cross-legged on a milk crate, studying a crayon-drawn map of the Henderson Community Center that she'd obtained through methods she refused to elaborate on. Empty coffee cups littered the floor around her, evidence of a fourteen-hour stakeout that had involved watching a bounce house inflate while pretending to be interested in a dog's birthday. Outside, the community center glowed with the innocent promise of juice boxes and sugar crashes, its parking lot currently occupied by three minivans and a sedan that definitely belonged to a grandmother. The van smelled of stale french fries, broken dreams, and the particular desperation of people who had chosen this life over dental school. Bill tapped a laminated photograph of seven-year-old Tyler Morrison's birthday party with a finger that had seen too many paper cuts. The image showed a bouncy castle, a dinosaur-shaped cake, and approximately forty wrapped presents arranged in a semicircle like offerings to a sugar god. "Forty presents, Molly. Forty. That's not a heist. That's a grocery run." Molly didn't look up from her map, her finger tracing the route past the supply closet. "It's forty opportunities. We go in during pin-the-tail. Parents distracted, kids blindfolded, presents sitting there like unwrapped gifts." "They're wrapped," Bill pointed out, his voice carrying the weight of a man who understood the physics of gift-giving. "That's the point of presents." "Camouflage," Molly said, finally meeting his eyes. "We blend in with the wrapping paper." Bill scratched his beard, which was currently decorated with a temporary butterfly tattoo that Molly had applied while he slept three nights ago. The ink had begun to fade into a watercolor smudge that made him look like a middle-aged man who had lost a bet at a baby shower. "What's our cover? I look like a man with opinions about craft beer." "We pose as entertainment," Molly said, pulling out a tutu that had seen better decades and smelled faintly of mothballs, regret, and the ghost of a previous clown who had not survived the profession. "I'm the clown. You're the traumatized assistant." "I love it," Bill said, holding up a rubber chicken with the enthusiasm of a man accepting a life sentence. "We make balloon animals, load the van during the song." "Precisely," Molly said, checking her watch with the intensity of a bomb defusal expert. "Hit the table during Happy Birthday. Classic misdirection." Bill adjusted his tactical vest over the tutu, creating a silhouette that defied several laws of fashion and physics. The combination of military hardware and ballet skirt made him look like a confused action figure that had been assembled by a committee of toddlers. "And if a parent asks why the quiet clown wears combat boots?" "Tell them it's for traction," Molly said, her voice dropping to a whisper. "Move, Bill. I want that dinosaur cake." Bill stood up, the tutu poofing around his knees like a failed soufflé, his combat boots squeaking against the linoleum. "Molly, these are children. Actual children with feelings. Therapy bills." "Think of it as wealth redistribution," Molly said, grabbing her duffel bag and checking that her lock picks were still hidden in her sock. "They have too many toys. Late-stage capitalism lesson." Bill sighed, a sound like a deflating whoopee cushion escaping into the void. "Lead the way, Clown Molly. If a kid asks why I'm crying, I'm blaming you." They stepped out into the night, two shadows dressed as a shared hallucination, ready to commit the most absurd felony in suburban history, their hearts hammering with the rhythm of impending felony and possibly felonious baked goods.
created_at
"2026-10-07T09:07:28.869Z"
10/7/2026, 9:07:28 AM
model
"mistralai/mistral-large-4-0-20261006"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2086
moderation_latency
(null)
generation_time
172599
tokens_prompt
75
tokens_completion
14269
native_tokens_prompt
67
native_tokens_completion
15424
native_tokens_completion_images
(null)
native_tokens_reasoning
13278
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
"default"
usage
0.03228172
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1791364048-pcQ5mBerFhpLRAxSzDGE"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1791364048-aPgzJLyaRGgP0otfJ6F0"
upstream_id
"93e49d606d35472290be45fcf8771a5c"
provider_responses
0
endpoint_id
"19b3b5d7-23f6-4b63-8109-64a6cc37d996"
id
"93e49d606d35472290be45fcf8771a5c"
is_byok
false
latency
920
model_permaslug
"mistralai/mistral-large-4-0-20261006"
provider_name
"Mistral"
status
200
total_cost
0.03228172
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mistral"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
640
17.72%
Dialogue to Total Word Ratio
Ratio: 24.12%, Deviation: 5.88%
neededClean
false
wordsTotal
651
wordsDialogue
157
8.8586%