NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-07-01KWZ49JTV43F54PVG5285X7KE
Run Details
Model:
Aion 3.0
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
18.9s
Cost:
$0.004905
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
6.64%
Dialogue to Total Word Ratio
(details)
3.3206%
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 sound of a toddler's birthday party filtered through the thin wall of the van, all squealing laughter and pop music and the distant wail of someone who definitely didn't get the first slice of cake. Molly pressed her binoculars against the window and let out a low whistle. "Sixteen kids," she said. "Maybe twenty. Hard to tell. They keep moving like a pack of feral hamsters." Bill didn't look up from the blueprint spread across his knees. He'd stolen it from the party rental company three days ago, along with a bouncy castle and an industrial-sized tub of frosting. The frosting had been an accident. The bouncy castle was for practice. "The presents are in the back room," he said, tapping the paper. "Right past the piƱata station. We wait for the magician, make our move during the big finale." "The magician." Molly lowered the binoculars. "You're telling me our entire heist hinges on a man in a top hat?" "His name is Magnificent Marco, and his big finish involves smoke pellets and a disappearing dove. I've seen his YouTube channel. It's terrible, but it's six minutes of pure chaos. More than enough time." Molly chewed her thumbnail. Through the window, she watched a small girl in a princess costume body-check a boy in a dinosaur onesie over a balloon animal. The dinosaur went down hard. The princess took the balloon sword and walked away without looking back. These were not the soft targets Bill had described in his pitch. Bill had promised soft targets. "What about the parents?" Molly asked. "Buffet table. Once the cake comes out, they're done. I've timed it. Cake comes out, Marco starts his act, every parent migrates to the snack table like salmon swimming upstream. The back room will be completely clear." "And if it's not?" Bill folded the blueprint with the care of a man who had once been a boy scout, briefly, before getting kicked out for selling the troop's equipment. He smiled. The smile was not reassuring. It was the smile of a man who had once tried to steal a police car because it was parked in a loading zone and he felt that was basically permission. "Then we improvise," he said. "You take the window. I take the door. We grab the loot, stuff it in the duffel bags, and we're out before anyone realizes what happened. The kids will be too hopped up on sugar to notice. The parents will be too busy arguing over who brought the stale hummus. We walk away clean." Molly looked at the duffel bags in the back of the van. They were large. They were very, very large. She had asked about this during the planning phase. Bill had said it was better to over-prepare. Molly had said it was better to not look like they were robbing a bank on their way to robbing a child's birthday party. They had compromised by getting duffel bags in a cheerful pink. Now they looked like they were robbing a bank and then going to a baby shower. "Fine," Molly said. "But if that dinosaur kid comes at me, you're on your own." Bill pulled on his ski mask. It had cartoon penguins on it. He'd bought it specifically for this job. "Fair enough."
created_at
"2026-07-07T20:28:43.756Z"
7/7/2026, 8:28:43 PM
model
"aion-labs/aion-3.0-20260707"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
691
moderation_latency
(null)
generation_time
18733
tokens_prompt
77
tokens_completion
885
native_tokens_prompt
67
native_tokens_completion
784
native_tokens_completion_images
(null)
native_tokens_reasoning
91
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
0.004905
router
(null)
provider_responses
0
endpoint_id
"8c8f827d-2bef-468a-8f9e-8e27b470de5e"
id
"528a84bb27d548c9b600897ab244b297"
is_byok
false
latency
691
model_permaslug
"aion-labs/aion-3.0-20260707"
provider_name
"AionLabs"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1783456123-nh0s6i2ffpw5SkRuu8Yu"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783456123-Ek1KANdVXqI9GxQ6n79q"
upstream_id
"528a84bb27d548c9b600897ab244b297"
total_cost
0.004905
cache_discount
(null)
upstream_inference_cost
0
provider_name
"AionLabs"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
551
6.64%
Dialogue to Total Word Ratio
Ratio: 37.36%, Deviation: 7.36%
neededClean
false
wordsTotal
554
wordsDialogue
207
3.3206%