NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY9RH2KG4A5SMYQ7FGHC3HW
Run Details
Model:
Ministral 3 14B
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
2.5s
Cost:
$0.000080
Provider:
Mistral
Evaluation summary
Score
Evaluator
Details Link
9.61%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
4.8041%
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
Molly adjusted her ski mask and grinned at Bill, who was already picking the lock on the back gate of the suburban mansion. "Alright, Bill, you got the gate, I got the presents. But we’re not taking *all* of them, right? Just the good stuff. The Lego sets, the action figures—" "Molly, we’re thieves, not charity workers. If it’s under a tree, it’s ours." She sighed. "Fine, but we skip the Barbie dolls. Too much emotional baggage." Bill finally popped the lock and pushed the gate open. "You’re the one who said we needed a ‘heart’ for this job." "Yeah, well, my heart’s in my stomach because we’re about to rob a *children’s party*." They crept through the backyard, dodging a trampoline and a half-eaten cupcake. Inside, the living room was packed with kids screaming over presents. "Okay, new plan," Molly whispered. "We grab the big stuff, then bolt before anyone notices." Bill nodded. "And if someone yells ‘thief’?" "Then we tell them Santa sent us." A little boy spotted them. "Mom! There’s two *very* suspicious adults!" Molly grabbed a giant teddy bear and a robot toy. "Run!" Bill snatched a remote-control car and a dollhouse. "We’re outta here!" They bolted for the gate as the party guests chased them, screaming. "Next time," Bill panted, "we hit a *baby shower*."
created_at
"2026-02-20T19:52:21.855Z"
2/20/2026, 7:52:21 PM
model
"mistralai/ministral-14b-2512"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
223
moderation_latency
(null)
generation_time
1920
tokens_prompt
77
tokens_completion
356
native_tokens_prompt
69
native_tokens_completion
333
native_tokens_completion_images
(null)
native_tokens_reasoning
0
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.0000804
router
(null)
provider_responses
0
endpoint_id
"0b53717b-9af5-4de3-80af-ebedd2adf981"
id
"20926ed999db4498a95d8958a04a4a61"
is_byok
false
latency
220
model_permaslug
"mistralai/ministral-14b-2512"
provider_name
"Mistral"
status
200
api_type
"completions"
id
"gen-1771617141-muw9W4NZcbkeZUZ4kuYN"
upstream_id
"20926ed999db4498a95d8958a04a4a61"
total_cost
0.0000804
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mistral"
Evaluation details
Result
Evaluator
Details
Meta Data
9.61%
Matches word count
n/a
neededClean
false
words
222
0.00%
Dialogue to Total Word Ratio
Ratio: 54.50%, Deviation: 35.50%
neededClean
false
wordsTotal
222
wordsDialogue
121
4.8041%