NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-27-01KQ79YAWFY0DNRP6PKD0VEV96
Run Details
Model:
Qwen3.6 Max Preview
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
3m 51s
Cost:
$0.067686
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
50.0000%
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 kitchen table vanished beneath a topographic map of a suburban cul de sac. Molly traced a highlighter across a laminated photograph of a vinyl bounce house. Bill adjusted the straps on a pair of oversized gardening kneepads. The fluorescent bulb overhead flickered with the rhythmic buzzing of a dying fly. Rain drummed against the windowpane in a steady, mocking rhythm. A half eaten donut sat near the blueprints, slowly hardening into a geological specimen. The radio crackled with static before falling completely silent. Shadows stretched across the linoleum as evening approached. Confirm the approach vector, Bill said. Do we scale the fence or use the side gate? Side gate, Molly replied. The fence has wind chimes. Acoustic tripwires. We enter at two pm during the sugar crash. Six minutes before cake. Six minutes to clear the gift table, Bill said. What about the entertainer? Molly slid a grainy photograph across the wood. The clown is neutralized. Trapped in the balloon industrial complex. He cannot stop twisting poodles. A psychological cage of his own making. Understood, Bill said. He tapped a pen against his clipboard. What about the pinata? Ignore the pinata, Molly replied. Tactical decoy. Last time yielded hard candy and regret. We want wrapped boxes. High value targets only. Large bows go in the duffel. Bill opened a velvet case containing lockpicks and a roll of scotch tape. Do we wear gloves, or will the static from the wrapping paper compromise us? Silk liners only, Molly said. She poured two cups of lukewarm coffee. And stop fidgeting. Remember the Henderson job? You triggered a musical greeting card and we had to sprint through a sprinkler system while a golden retriever judged our life choices. That dog had no moral high ground, Bill muttered. He checked his wristwatch. What is the extraction route? Through the azaleas, past the inflatable castle, into the drainage culvert behind the compost bin, Molly said. I already greased the manhole hinges. It squeaks, so we will time the lift with the chorus of Happy Birthday. Acoustic cover is everything. You really planned for the compost, Bill said. He zipped his tactical vest. I always plan for the compost, Molly replied. Now check the inventory. Do we have the box cutters, the silent tape, and the juice pouches? Juice pouches? Bill frowned at the cardboard box on the floor. For blending in, Molly said. If a toddler spots you, hand him apple juice and keep moving. It is the ultimate distraction. Do not make eye contact. Do not accept finger paintings. We are professionals. Copy that, Bill said. Pouches loaded. Liners ready. I will bring the reinforced bag. Perfect, Molly said. She killed the overhead light. We hit the table, we bag the loot, we vanish before the candles are lit. This is not a birthday party. This is a tactical acquisition of overpriced electronics. Stay sharp. Stay quiet. And do not pop a single balloon. I would never, Bill said. I respect the latex.
created_at
"2026-04-27T11:07:11.127Z"
4/27/2026, 11:07:11 AM
model
"qwen/qwen3.6-max-preview-20260420"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2320
moderation_latency
(null)
generation_time
230999
tokens_prompt
77
tokens_completion
8544
native_tokens_prompt
76
native_tokens_completion
8665
native_tokens_completion_images
(null)
native_tokens_reasoning
7987
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"
usage
0.0676858
router
(null)
provider_responses
0
endpoint_id
"ca7fc0c7-f23d-4ac5-b2e0-945be5b61787"
id
"chatcmpl-b0158a87-0ac6-9ab1-a660-62fda8a2fb4f"
is_byok
false
latency
2319
model_permaslug
"qwen/qwen3.6-max-preview-20260420"
provider_name
"Alibaba"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777288031-gw8CrYpLtsRLRGJ5weeO"
session_id
(null)
api_type
"completions"
id
"gen-1777288031-ADNmK6Y46zSbldKohoLb"
upstream_id
"chatcmpl-b0158a87-0ac6-9ab1-a660-62fda8a2fb4f"
total_cost
0.0676858
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Alibaba"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 70.00%
neededClean
false
wordsTotal
500
wordsDialogue
0
50.0000%