NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-24-01M39X6RSJ8WY7KTRAD92T7VMD
Run Details
Model:
Aion 3.5 Mini (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 53s
Cost:
$0.007845
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
3.46%
Dialogue to Total Word Ratio
(details)
1.7318%
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
The plan was spread across the card table like a crime scene: blueprints of a suburban backyard, a map of a residential neighborhood, and a diagram of a bounce house that Molly had drawn herself, complete with little stick figures fleeing the scene. A red foam nose sat on the corner of the blueprints like a bloody period at the end of a very stupid sentence. Bill stared at it for a long time. Then he went back to the diagram, hoping it would have changed. "You want to rob a birthday party." "Liberate a birthday party. There's a difference." "Those sound like the same thing." "Moral relativism, Bill. Keep up." Molly tapped the diagram, leaving little ink dots on the bounce house. "The target is Timmy Henderson's seventh birthday. Saturday, two p.m. Presents on a folding table next to the bounce house. Parents distracted by cake. Kids contained in the inflatable structure." "So we steal from the children while they're bouncing." "While they're bouncing, yes. It's the perfect crime. No security cameras. No guard dogs. No witnesses who can legally drive." "One of them could testify. Timmy was a witness at the museum job." "That was a coincidence and you know it." "He pointed at you and said 'that's the lady from the painting.' We got lucky his mom thought he meant the Mona Lisa." "Which is why this time, we blend in." Molly pulled the clown costume from her bag and unfurled it across the table like the flag of a nation nobody wanted to visit. "I do balloon animals. You run the gift table." "I don't know how to do balloon animals." "You don't need to. Twist the balloon until it looks like a dog. Crying child? Hand it a dog. Crying harder? Second dog. Foolproof." "And the presents?" "Empty gift bag. Oversized coat. Classic misdirection." She folded the costume with the tenderness of a mother tucking in a child. "By the time Timmy unwraps his haul and finds nothing but paper and a card from his grandmother, we'll be three states away." "His grandmother's card will be there?" "I'm not a monster, Bill. The card stays." Bill opened his mouth, thought better of it, and closed it again. It was hard to argue with a person who had a moral code, even if the code was "steal everything except greeting cards." "What happens if a parent asks why the clown is loitering by the presents?" "You tell them you're the present coordinator. Clowns have job titles now. It's a modern industry." "What if Timmy recognizes me? He's seen my face. He pointed at it." "Then you wear the nose." Molly slid the foam nose across the table, where it came to rest against his coffee mug like a tiny red ship docking. "Nobody recognizes a man with a red foam nose. It's science." "It's not science." "It's clown science, which is stronger, because nobody studies it." Bill looked at the diagram one last time, at the little stick figures bouncing joyfully, oblivious to their future disappointment. Somewhere in a nice suburban backyard, a seven-year-old was unaware his entire year was about to be ruined by two adults in face paint. "Fine," he said. "But if this goes wrong, I'm telling the police you made me do it." "You'll be in a clown suit, Bill. They'll never believe a word you say."
created_at
"2026-09-24T14:30:41.22Z"
9/24/2026, 2:30:41 PM
model
"aion-labs/aion-3.5-mini-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1809
moderation_latency
(null)
generation_time
113220
tokens_prompt
75
tokens_completion
5956
native_tokens_prompt
65
native_tokens_completion
5595
native_tokens_completion_images
(null)
native_tokens_reasoning
5137
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
0.0078447
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790260241-ghzLT9w5AxZ8O5Et1ym2"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790260241-OLmsN3kf9O7Ck1F0P0bC"
upstream_id
"b68a2aef39eb49bcaa051dd26fb975b6"
provider_responses
0
endpoint_id
"60ac3bdc-90ec-4769-9193-2c00b4d79562"
id
"b68a2aef39eb49bcaa051dd26fb975b6"
is_byok
false
latency
1809
model_permaslug
"aion-labs/aion-3.5-mini-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.0078447
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
0.00%
Matches word count
n/a
neededClean
false
words
565
3.46%
Dialogue to Total Word Ratio
Ratio: 58.20%, Deviation: 8.20%
neededClean
false
wordsTotal
567
wordsDialogue
330
1.7318%