NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-13-01M4TBX8E406BQG0BQT8DRNSMZ
Run Details
Model:
Ember-1 (Reasoning, Max)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
2m 39s
Cost:
$0.131657
Provider:
Fireworks
Evaluation summary
Score
Evaluator
Details Link
27.17%
Matches word count
(details)
89.37%
Dialogue to Total Word Ratio
(details)
58.2691%
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 unrolled the blueprint across the hood of their rusted sedan, weighing down the corners with a half-eaten gas station burrito. Across the street, balloons sagged on the target's mailbox beneath a glittery sign: HAPPY 6TH BIRTHDAY TYLER. "This is the target," she said. "Six-year-old's birthday. Backyard. Bouncy castle here, snack table here, and here—" she tapped a rectangle drawn in crayon, "—the present table. Mother lode." Bill squinted, orange cheese-puff dust on his fingers. "Why's the blueprint in crayon?" "I stole it off the kid's refrigerator door. It's called reconnaissance, Bill." "Why's there a dog on it?" "That's a pony. They're getting a pony." Bill whistled. "Are we stealing the pony?" "We're stealing the presents, Bill. Focus." "Could we steal the pony?" "No." "Only, my cousin's got a trailer—" "Bill." "Right. Presents." A mother pushing a stroller walked past. They waved like normal citizens. She waved back. Molly waited until she turned the corner. "Party starts at two. At two-thirty, the clown arrives. That's our window." "The clown's our window?" "Nobody watches anything when a clown's working. Kids are mesmerized. Parents check their phones. The clown is a walking distraction." "What if the clown sees us?" "What's he gonna do, honk at us?" Bill considered this. "My Uncle Ray was a clown. He'd have honked." "Your Uncle Ray was arrested at a mini-golf course, so maybe we don't use him as the benchmark." "That was a misunderstanding about a windmill." Molly pinched the bridge of her nose and jabbed a finger at the blueprint. "We go in dressed as the balloon animal guy's assistants. I told Tyler's mom we're with the party company." "There's a balloon animal guy?" "There is now. You're him." "I can't make balloon animals." "You don't have to. You just have to look like a man who regrets his life choices, which, frankly, is your best look." Bill frowned. "What's your job?" "I'll be by the present table with a big gift bag, 'reorganizing.' Everything goes in the bag. Out the side gate. Into the car. Ninety seconds, tops." "And what do we do if a kid catches us?" "You make him a balloon animal." "I can't make balloon animals, Molly!" "Make a balloon snake. It's the same balloon, you just don't do anything. Tell him it's a worm. Kids love worms." Bill opened his mouth to argue, then closed it. She was right. Kids did love worms. "What's the take? What's in a six-year-old's presents?" Molly smiled like a woman who'd done the math. "Legos, Bill. Untraceable. High resale value. And at least one grandma always gives a savings bond." "How much is a savings bond worth?" "Fifty bucks." "We're robbing a child for fifty bucks?" "Plus Legos." Bill thought about it, then nodded slowly. "When do we leave?" "Saturday. Wear something festive." "Like a tie?" "Like a man who makes balloon animals for money and hates himself." "I've got just the shirt." Molly rolled up the blueprint, grabbed the burrito, and headed for the driver's seat. Behind them, somewhere in the target's backyard, a pony neighed, and Bill looked back longingly.
created_at
"2026-10-13T10:11:10.926Z"
10/13/2026, 10:11:10 AM
model
"fireworks/ember-1-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1396
moderation_latency
(null)
generation_time
158881
tokens_prompt
75
tokens_completion
7931
native_tokens_prompt
149
native_tokens_completion
8774
native_tokens_completion_images
(null)
native_tokens_reasoning
7931
native_tokens_cached
148
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
0.1316574
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1791886270-2yK4fPBfBFnA3lBfX7YK"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1791886270-GTLg65GmMxaUhbUXe772"
upstream_id
"chatcmpl-3f027c29a7564277a4ff528c0ce27b72"
provider_responses
0
endpoint_id
"b89a2f05-4051-4446-a9b0-7bc8ae7b7c90"
id
"chatcmpl-3f027c29a7564277a4ff528c0ce27b72"
is_byok
false
latency
1396
model_permaslug
"fireworks/ember-1-20260923"
provider_name
"Fireworks"
status
200
total_cost
0.1316574
cache_discount
0.0003996
upstream_inference_cost
0
provider_name
"Fireworks"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
27.17%
Matches word count
n/a
neededClean
false
words
519
89.37%
Dialogue to Total Word Ratio
Ratio: 68.50%, Deviation: 1.50%
neededClean
false
wordsTotal
527
wordsDialogue
361
58.2691%