NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-11-01KKFK657AF7BP0S6VV0DHD6Y3
Run Details
Model:
Qwen 3.5 9B
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
2m 13s
Cost:
$0.001519
Provider:
Together
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.09%
Dialogue to Total Word Ratio
(details)
0.0464%
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 humid air of the suburban cul-de-sac clung to Molly's expensive trench coat. She adjusted her grip on a stack of heavy wrapping paper and looked at Bill, who was currently polishing a delayed birthday balloon on the ground. "Are you ready to go through with this robbery tonight?" she asked, her eyes darting nervously toward the house ahead where the music played loud and relentless. "It feels wrong, Bill." Bill didn't blink. He just kept polishing the balloon. "Nothing is wrong, just different," Bill whispered, dangling the red balloon in front of his face. "We hit the trays. It is a food fight of sorts." "You mean the crowns." "The crowns are empty, Molly. It is a toy war underneath." "But they have guns." "Plastic guns are good for nothing but distractions. It is a distraction to the adults." "What about the cake?" "We avoid the cake. That is for the insecure parents. You cannot have the cake." "I hope you know what you are talking about." "I know all about it." "Then tell me how we get inside this house." "We climb the shrubbery. It is secret. It is the garden way." "But the dogs." "The dogs are inside the house. They are watching the TV screen." "The dogs are loud." "They are only barking for food. Just the thrill." "Okay, what about the window?" "That is dangerous. It is glass. It breaks." "Can we jump?" "We can't jump. We are not parkour. We are climbers." "Where do we go from here?" "We go to the left." "Left. Okay." "No. Left. No." "No, no." "Left then." "We are confused." "We are fine now. We are locked." "Okay. We go. Go forward." "Forward." "Wait. I see the balloon." "The balloon holds helium." "Do we blow it up?" "No. We smash it." "The balloon is red." "Smash it red." "Okay. Smash it." "We have done the first thing." "Yes. It is done." "Oh, it is done." "Look at the bag." "Bag." "Put the bag in the bag." "Put the bag inside." "We need space." "We have space now. It is fine." "We need to move." "We are moving. We are going." "Wait. Is it done?" "It is done." "Okay. Go." "Let's go. Now." "We go now." "Yes. Go." "Molly, stop now." "Stop." "Stop?" "No. Yes, stop." "Okay. We stop." "Wait. What?" "I stop now." "You stop." "I stop." "Then stop." "Okay. Stop." "Wait. Let's go again." "No. No more go." "I said no." "No more." "Okay. We stop." "Wait. Wait, why?" "No reason." "Okay. I stop." "Good."
created_at
"2026-03-11T23:19:52.823Z"
3/11/2026, 11:19:52 PM
model
"qwen/qwen3.5-9b-20260310"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
332
moderation_latency
(null)
generation_time
132594
tokens_prompt
77
tokens_completion
9473
native_tokens_prompt
76
native_tokens_completion
10077
native_tokens_completion_images
(null)
native_tokens_reasoning
8758
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
0.00151915
router
(null)
provider_responses
0
endpoint_id
"00d506d9-a5e9-4382-a2b9-d54732365594"
id
"oaguNpe-3pDw3Z-9dae495b33a02d57"
is_byok
false
latency
332
model_permaslug
"qwen/qwen3.5-9b-20260310"
provider_name
"Together"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1773271192-sudRAahRJe2vgppB06WS"
upstream_id
"oaguNpe-3pDw3Z-9dae495b33a02d57"
total_cost
0.00151915
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Together"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
427
0.09%
Dialogue to Total Word Ratio
Ratio: 81.82%, Deviation: 11.82%
neededClean
false
wordsTotal
429
wordsDialogue
351
0.0464%