NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-07-01KWZ4MYC5BB6YGX4PK3VTX4V4
Run Details
Model:
Aion 3.0 Mini
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 10s
Cost:
$0.004088
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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 the focus on her binoculars, peering through the slats of the neighbor’s fence. A dozen children in party hats were destroying a piñata shaped like a unicorn, and the screams of glee were almost feral. “Okay,” she whispered to Bill. “We go in during the magician.” Bill shifted his weight, his tool belt jingling with lockpicks and a can of Silly String. “But the magician’s a grown-up. Grown-ups can call the cops.” “He’s also sixty and wears a velvet cape. He is not a threat.” She pointed a gloved finger. “The presents are under that table. The big box is the main target. Pink wrapping with a bow the size of a soufflé.” “That’s for the birthday girl,” Bill said. “Her name is Chloe. She is six. She likes ponies.” Molly lowered the binoculars. “Do you have the list of weaknesses?” Bill consulted a crumpled notebook, licking his thumb to turn a page. “Bouncy castle operator—easily distracted. Milks the slide with soap. Weakness: thermal detonator. Birthday mom—anxiety-prone. Might cry. Weakness: missing sequins on the tablecloth. Grandmother—glaucoma and still has a mean left hook. Weakness: cats.” “What about the magician?” Bill flipped two pages. “The Amazing Roberto. Real name: Kevin. Major weakness: his doves are rented, and he is terrified of anything that might make them escape. If one dove goes rogue, he will chase it for fifteen minutes. He also can’t do a card trick without sweating through his cummerbund.” Molly smiled. “Perfect. We use the dove trick. I’ll pop the cage. You grab the haul. We meet at the van. No eye contact. No stopping for cake.” Bill hesitated. “Molls… it is a kid’s birthday. Her grandmother knitted her that sweater. It has a dog on it. What if she cries?” “I have a plan for crying children, Bill. It is called a five-second delay and a firm ‘look, a fire truck.’ If that fails, I deploy a decoy balloon animal. Now, are you in or out?” “In,” he sighed. “But I am not using the thermal detonator on the bouncy castle. That is a war crime. The Geneva Convention specifically bans booby-trapping inflatables.” “Fine.” Molly pulled out a small metal sphere from her pocket. “We will use the glitter bomb instead. It is less explosive, but it sends up a cloud of shimmer that will blind everyone for exactly seven seconds. It creates a perfect window.” “But glitter gets everywhere,” Bill said. “It is the herpes of craft supplies. We will be finding sparkly blue flecks in our socks for years.” “Exactly. It is our signature. People will speak of a Glitter Bandit.” She checked her watch. “Showtime. Get the dove leash.” They crouch-crawled toward the fence, but Bill stopped at a loose board. “Molly, one more thing. What if the present is… socks?” “Then we fence the socks. There is a massive black market for high-quality cotton crew socks. They go for twenty a pair unopened.” “And if it is a pony?” “Ride it out.” Bill grinned, for the first time in the entire stakeout. “Okay. Let us steal a pony for a crying six-year-old. I can live with that.” Molly rolled her eyes. “Priorities, Bill. We are *professionals*. We are not stealing a pony. That is a living creature. You cannot fence a pony on the black market.” “Can we at least take a photo with the pony?” “After the heist, Bill. Focus.” They slipped through the fence.
created_at
"2026-07-07T20:34:56.027Z"
7/7/2026, 8:34:56 PM
model
"aion-labs/aion-3.0-mini-20260707"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1721
moderation_latency
(null)
generation_time
69653
tokens_prompt
77
tokens_completion
3189
native_tokens_prompt
66
native_tokens_completion
2887
native_tokens_completion_images
(null)
native_tokens_reasoning
2266
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
0.004088
router
(null)
provider_responses
0
endpoint_id
"0ad7ab68-a351-4a96-b2ef-fa8fe13adae3"
id
"fd347d88227c4283a4193836e5358a7e"
is_byok
false
latency
1721
model_permaslug
"aion-labs/aion-3.0-mini-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-1783456496-yZPPB7iaEQRZIlUCCkGC"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783456496-3umzd6GHZPIZ2NYHs8YO"
upstream_id
"fd347d88227c4283a4193836e5358a7e"
total_cost
0.004088
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
578
0.00%
Dialogue to Total Word Ratio
Ratio: 76.45%, Deviation: 26.45%
neededClean
false
wordsTotal
586
wordsDialogue
448
0.0000%