NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHK3D26VJCW0N205VCZHA3FC
Run Details
Model:
Qwen 3.5 397B A17B
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
5m 59s
Cost:
$0.049654
Provider:
Alibaba
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 spread the laminated floor plans across the sticky dashboard of the getaway van. The interior smelled of stale coffee and impending regret. Outside, the community center glowed with festive streamers and the innocent laughter of children who had no idea their financial future was being liquidated by career criminals. Bill chewed on the end of an unlit cigar, his brow furrowed deeply as he studied the detailed diagram of the gift table. The van engine ticked as it cooled down in the night air. Shadows lengthened across the pavement. This is really low, even for us, Bill said quietly. We usually hit banks. It is not low, Molly replied sharply, tapping a red circle around the punch bowl. It is high yield with minimal security resistance. These people do not expect professional extraction teams at a birthday bash. Stop complaining. There are clowns, Bill noted, pointing a thick finger at the sketch. Actual clowns. They know tricks. They are unpredictable agents of chaos. Clowns are chaotic variables, nothing more. We move during the cake cutting. The sensory overload of singing provides perfect acoustic cover. Molly pulled a pair of binoculars from the glove box and scanned the entrance. Two parents stood guard near the coat rack, but their attention was fixed on a toddler attempting to eat a balloon. She adjusted her focus ring carefully. The coast was clear for now. What about the parents? Bill asked. They are fierce when their young are involved. Like bears. They are distracted by sugar and social validation. We wear the uniforms. Molly tossed a polyester vest onto Bill's lap. It read STAFF in cheerful yellow letters. The fabric felt cheap and synthetic against his skin. He examined the stitching with suspicion. You are in charge of plates. I am handling gifts. Do not drop anything. Bill held the vest like it was radioactive. He looked at the community center windows. Warm light spilled out onto the pavement. A shadow moved behind the glass. I do not feel good about this, Molly. These are six-year-olds. They believe in magic. It is wrong. They believe in plastic swords and action figures. We are simply redistributing wealth. Take the good stuff. Molly checked her watch. The window is opening. The magician is starting his act. That means hypnosis and scarves. Perfect distraction. If we get caught, I am telling them you made me do it, Bill grumbled. pulling the vest over his head. It smelled of spilled juice and despair. He zipped up his jacket to hide the uniform. Just load the bags. Focus on the envelopes with cash. Ignore the homemade cards. Move fast. Molly opened the side door and slipped into the shadows of the parking lot. Bill followed, his heavy boots crunching on gravel. He looked like a gopher wearing a party vest. He waddled slightly. The night air was cold against his face. Inside, the sound of happy birthday echoed off the linoleum. Molly signaled near the pile of wrapped boxes. Bill hesitated, looking at a small boy blowing out candles. He sighed, hoisted a duffel bag, and began shoving LEGO sets into the darkness. It was the easiest score of their careers, but somehow the heaviest on their souls. The guilt was tangible.
created_at
"2026-02-16T11:35:34.792035+00:00"
2/16/2026, 11:35:34 AM
model
"qwen/qwen3.5-397b-a17b-20260216"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
841
moderation_latency
(null)
generation_time
358463
tokens_prompt
77
tokens_completion
12582
native_tokens_prompt
76
native_tokens_completion
13780
native_tokens_completion_images
(null)
native_tokens_reasoning
13081
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.0496536
router
(null)
provider_responses
0
id
"chatcmpl-e0120104-9844-99f1-be00-29227e48d54c"
status
200
is_byok
false
latency
841
endpoint_id
"180484a2-e24b-495a-ba41-b34b680d5a9f"
provider_name
"Alibaba"
model_permaslug
"qwen/qwen3.5-397b-a17b-20260216"
api_type
"completions"
id
"gen-1771241375-qNzdaGftePz4IiMUrYGP"
upstream_id
"chatcmpl-e0120104-9844-99f1-be00-29227e48d54c"
total_cost
0.0496536
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Alibaba"
Evaluation details
Result
Evaluator
Details
Meta Data
0.0000%
Matches word count
n/a
neededClean
false
words
546
0.0000%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 30.00%
neededClean
false
wordsTotal
546
wordsDialogue
0
0.0000%