NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-13-01M4TC3RM7SWHXA1T4J5XNBMY9
Run Details
Model:
Ember-1 (Reasoning, Max)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
2m 15s
Cost:
$0.150972
Provider:
Fireworks
Evaluation summary
Score
Evaluator
Details Link
60.28%
Matches word count
(details)
50.09%
Dialogue to Total Word Ratio
(details)
55.1805%
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 unrolled the blueprint across the hood of the car and weighed down the corners with a juice box, a slap bracelet, and a small plush giraffe. The whole thing was drawn in crayon, because the only person who could verify the Hendersons' floor plan was seven years old and charged by the fruit snack. Red yarn connected the back door to the gift table, and a green blob in the yard was labeled, in Molly's careful handwriting, DO NOT TRUST THE DOG. "Saturday, two p.m.," she said. "Forty to fifty wrapped presents, right here on the gift table." Bill squinted at the diagram. "What's this squiggly bit?" "Bouncy castle. Both an obstacle and an opportunity." "How is it an opportunity?" "Nobody guards gifts mid-bounce, Bill." He nodded slowly, the way he did when something was either brilliant or insane, which with Molly was usually both. They had robbed a casino, a yacht, and once, memorably, a locksmith's convention, but this was their first job where the security detail napped after lunch. Bill produced a notepad, because lists made crime feel responsible. "What's our cover?" "Clowns." "No." "Magicians?" "You lost my wallet doing magic." "Caterers, then. Nobody questions a caterer." Bill wrote it down in small, neat letters. "Getaway?" "Gerald's ice cream truck. Fifteen percent." "Fifteen? It plays Turkey in the Straw at half speed." "It was that or a hot air balloon, and you get nosebleeds." Gerald had a record, a chest freezer, and no questions, which made him practically family. Bill conceded the point with a grunt. Across the park, a jogger glanced at the crayon blueprint, the giraffe, and the two grown adults crouched over the hood like conspirators in a very low-stakes war room, and wisely kept jogging. Molly capped her crayon and fixed Bill with the look she reserved for jobs that were either foolproof or catastrophically stupid. History suggested it could be both before lunch. "One rule," she said. "We take the presents, not the cake." "Why not the cake? I like cake." "Stealing a kid's cake is how you end up in a documentary. We're thieves, Bill, not monsters." This was, Bill knew, the foundation of their entire professional reputation. Lines mattered. Lines kept you lovable. Cross them, and suddenly you were the villains in somebody's therapy session. "The kid's name is Tyler," Molly added. "He's turning eight." Bill checked his watch. "Forty-eight hours. You think we're ready?" Molly surveyed the crayon blueprint, the giraffe paperweight, and the man beside her who had once been arrested mid-juggle at a birthday party he wasn't even robbing. "Ready as ever. Grab the gifts, smile at the parents, and if anyone asks, we're with the balloon company." "There's no balloon company." "There is now. I had cards made." Bill shook his head, but he was smiling as he tucked the notepad away. Somewhere across town, a small mountain of presents sat wrapped and waiting in shiny paper, and come Saturday afternoon, Tyler Henderson was going to have a very memorable eighth birthday. Just not for the reasons his parents had planned.
created_at
"2026-10-13T10:14:44.115Z"
10/13/2026, 10:14:44 AM
model
"fireworks/ember-1-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
4516
moderation_latency
(null)
generation_time
134847
tokens_prompt
75
tokens_completion
8299
native_tokens_prompt
149
native_tokens_completion
10035
native_tokens_completion_images
(null)
native_tokens_reasoning
8299
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
0.150972
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1791886484-VbgIjya7rTe0pynl8FVC"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1791886484-5qAXIWrv59iznTDmnFdV"
upstream_id
"chatcmpl-ab30246e312f45f0b6a9da541ce09532"
provider_responses
0
endpoint_id
"b89a2f05-4051-4446-a9b0-7bc8ae7b7c90"
id
"chatcmpl-ab30246e312f45f0b6a9da541ce09532"
is_byok
false
latency
4516
model_permaslug
"fireworks/ember-1-20260923"
provider_name
"Fireworks"
status
200
total_cost
0.150972
cache_discount
(null)
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
60.28%
Matches word count
n/a
neededClean
false
words
515
50.09%
Dialogue to Total Word Ratio
Ratio: 33.72%, Deviation: 3.72%
neededClean
false
wordsTotal
519
wordsDialogue
175
55.1805%