NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-24-01M39WDB8SD354181J8JTK7XD5
Run Details
Model:
Aion 3.5 Mini (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 33s
Cost:
$0.005420
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
90.48%
Matches word count
(details)
2.20%
Dialogue to Total Word Ratio
(details)
46.3431%
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
The warehouse smelled like old pizza and stolen dreams. Rain ticked against the skylight. Molly slapped a rolled-up blueprint onto the crate they used as a table, and Bill flinched like it was a live grenade. "Okay," Molly said. "The Henderson kid turns six on Saturday. Two hundred guests, one rented bouncy castle, and per my source, an estimated forty pounds of wrapped presents." Bill rubbed his chin. "Forty pounds. That's a lot of Lego." "It's a lot of everything. Plastic junk worth maybe four hundred bucks retail, but the hot toys this season? Triple on the gray market." "We're stealing from a six-year-old." "We're redistributing joy." "His joy." "Joy has no owner, Bill. It's on a mug in my kitchen." Bill picked up the blueprint, squinted at it, then turned it upside down. "Also, last month we robbed a garden gnome collector. Before that, a beehive museum. Now this. We're hitting an all-time low, Molly." "We're hitting an all-time adorable. There's a difference. Now focus. Which one of these blobby shapes is the house?" "The blobby shape is the bounce castle. The house is the rectangle. Security is one attendant named Doug. Ex-military. Taser. Whistle." "A whistle?" "He's very serious about the whistle. Kids have been escorted out for double-bouncing. Adults, too." "So how do we get past Doug?" "We don't get past Doug. We perform for Doug. I booked us as The Amazing Marvo and his fabulous assistant, Wanda." Bill went very still. "Why am I Wanda?" "Because 'Marvo' was taken, and you lost the coin toss." "I lost the coin toss last time. I'm always Wanda." "And yet the wig suits you." Bill sighed and pulled a sequined glove from his jacket. "Fine. What's the play?" "You do the magic show, keep forty kids hypnotized, and I work the present table by the fence. Swap bags, fill a duffel, out through the hedge. Twelve minutes, tops." "What about the piñata?" "The piñata is a decoy. Allergy-safe candy and coupons for a car wash. Ignore it." "And the pony?" Molly paused. "What pony?" "The invitation says there's a pony. If there's a pony, I'm petting the pony. That's non-negotiable." "Fine. Pet the pony. Between the show and the loot grab. Quickly. And remember, the getaway van is parked out front disguised as an ice cream truck." "If I get hungry mid-job—" "Don't eat the inventory." "That's cold, Molly." "The inventory is frozen yogurt. It stays cold. That's the point. Now, what if the kid cries?" "Kids cry at their own parties, Bill. Nobody notices one more." Bill nodded slowly, then held up a finger. "One question. The birthday card. The big one, shaped like a dinosaur, signed by everyone. We can't take that. It's sentimental." Molly stared at him for a long moment. "What? I have a heart." "You have a heart. Fine. We leave the card." "Thank you." "But we're taking the pony." "WHAT?" "Kidding." Molly rolled up the blueprint and grinned. "Mostly." Across town, a six-year-old slept soundly, unaware his dinosaur cake was already spoken for.
created_at
"2026-09-24T14:16:48.166Z"
9/24/2026, 2:16:48 PM
model
"aion-labs/aion-3.5-mini-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2249
moderation_latency
(null)
generation_time
93394
tokens_prompt
75
tokens_completion
3662
native_tokens_prompt
65
native_tokens_completion
3863
native_tokens_completion_images
(null)
native_tokens_reasoning
2884
native_tokens_cached
65
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
0.0054199
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790259408-ksQNdOkifRB5dP30z7TS"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790259408-lKUEzHckvSOguT6ZqzRv"
upstream_id
"e2a9665599a248eeb26a7d28173d51cb"
provider_responses
0
endpoint_id
"60ac3bdc-90ec-4769-9193-2c00b4d79562"
id
"e2a9665599a248eeb26a7d28173d51cb"
is_byok
false
latency
2236
model_permaslug
"aion-labs/aion-3.5-mini-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.0054199
cache_discount
0.0000338
upstream_inference_cost
0
provider_name
"AionLabs"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
90.48%
Matches word count
n/a
neededClean
false
words
510
2.20%
Dialogue to Total Word Ratio
Ratio: 78.74%, Deviation: 8.74%
neededClean
false
wordsTotal
522
wordsDialogue
411
46.3431%