NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-17-01KHNEVAQJFERAYV9RGY0DZYQC
Run Details
Model:
Minimax M2.5
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
16.8s
Cost:
$0.000880
Provider:
Fireworks
Evaluation summary
Score
Evaluator
Details Link
51.93%
Matches word count
(details)
12.52%
Dialogue to Total Word Ratio
(details)
32.2235%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 crouched behind the inflatable bouncy castle, her black mask slipping down her nose. "I still think this is insane." Bill adjusted his clown wig, which smelled aggressively of synthetic strawberry. "You're the one who said we needed a fresh start. What's fresher than a seven-year-old's birthday party?" "The FBI isn't looking for us at a seven-year-old's birthday party," Molly whispered, watching a cluster of children sprint past clutching juice boxes. "They're looking for us at banks. At museums. At places that actually have stuff worth stealing." "But they won't expect this." Bill pulled out a crumpled diagram he'd drawn on a napkin. "Here's the plan. The party's at the Hendersons' house. Three entrances. Kitchen, living room, and that sliding door that goes to the backyard." "The backyard where all the presents are currently stacked under a banner that says 'Happy Birthday Emma'?" Molly raised an eyebrow. "Bill, those are toys. Plastic toys meant for children." "Toys that cost eighty dollars each." Bill tapped the napkin. "That pile is worth at least two thousand dollars. Do you know how many burritos two thousand dollars buys?" "I'm concerned about how quickly you did that math." "Four hundred burritos, Molly. Four hundred." Molly sighed, rubbing her temples. "Okay, let's say we somehow get past the parents. There's going to be a clown. There's going to be a face painter. There's going to be at least twenty children screaming in matching party hats." "That's the beauty of it." Bill grinned. "We become part of the chaos. I sign us up as entertainment." "You signed us up as entertainment?" "Face painter and assistant. I got us matching smocks and everything." He held up two small plastic aprons covered in rainbow polka dots. "They'll fit right over our gear." Molly stared at the smocks. "You want us to rob a children's party while dressed as face painters." "It's genius. Who's going to suspect the friendly face painting duo?" "The parents. The parents are going to suspect us." "The parents are going to be three mimosas deep watching their kids run around a bouncy castle. They'll probably thank us for keeping the little ones occupied while we quietly transfer merchandise to our bags." "Merchandise," Molly laughed despite herself. "You mean stolen toys." "Repurposed assets." Bill checked his watch. "We have four hours to prep. I got us fake IDs under the names of Sandra and Denise. Very normal face painter names." "Sandra and Denise." "Very trustworthy." Molly looked at the smocks, then at the distant sound of children's laughter, then at her partner's enthusiastic clown-wigged face. "If we get arrested for stealing from a seven-year-old, I'm blaming you entirely." "Fair. Now let's go practice our smiley faces." Bill stood up and stretched. "We need to look like people who love children." "I hate children." "Then you're going to need to practice your acting."
created_at
"2026-02-17T09:28:23.798609+00:00"
2/17/2026, 9:28:23 AM
model
"minimax/minimax-m2.5-20260211"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5930
moderation_latency
(null)
generation_time
16704
tokens_prompt
77
tokens_completion
875
native_tokens_prompt
97
native_tokens_completion
709
native_tokens_completion_images
(null)
native_tokens_reasoning
105
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
0.0008799
router
(null)
provider_responses
0
id
"58397f26-c0bc-429e-8287-7cf0218b7279"
status
200
is_byok
false
latency
5926
endpoint_id
"c61d37ec-745c-4efd-8129-a68371e056b2"
provider_name
"Fireworks"
model_permaslug
"minimax/minimax-m2.5-20260211"
api_type
"completions"
id
"gen-1771320486-Tn9usfQyWqgbqoXuXamB"
upstream_id
"58397f26-c0bc-429e-8287-7cf0218b7279"
total_cost
0.0008799
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Fireworks"
Evaluation details
Result
Evaluator
Details
Meta Data
51.93%
Matches word count
n/a
neededClean
false
words
484
12.52%
Dialogue to Total Word Ratio
Ratio: 76.45%, Deviation: 6.45%
neededClean
false
wordsTotal
484
wordsDialogue
370
32.2235%