NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-10-01KR912NEJ05GCQW3Q6301YSZJ
Run Details
Model:
Gemma 4 31B (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 8s
Cost:
$0.000934
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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
Molly spread the napkin across the dashboard of the sedan. Look at the blueprints, Bill. The target is a six-year-old's birthday bash. Bill squinted at the drawing. Blueprints? You drew this on a grease-stained napkin from the diner. It is a tactical sketch, Molly snapped. The cake is the epicenter. We strike during the singing of Happy Birthday. We are stealing toys, Molly. Not the Crown Jewels. Don't be naive, she whispered. Do you know what a mint-condition vintage holographic Charizard goes for on the black market? I thought it was just a party, Bill sighed. It is a goldmine of plastic and plush! Now, the infiltration. I will enter through the dog door. You are five-foot-eight. You are not fitting through a dog door. I have been stretching, Molly insisted. You will handle the diversion. What diversion? You will trigger the piƱata, she explained, her eyes gleaming. While the children are in a sugar-induced frenzy fighting over candy, we swoop in, bag the presents, and vanish into the hedges. The hedges are three feet tall, Bill pointed out. Tactical crouching, Bill! Get with the program. And what about the parents? There will be adults there. Distracted by mimosas and the sheer horror of twenty toddlers in one room, Molly said. They will not see a thing. What if someone sees us? Then we deploy the smoke bombs. Smoke bombs at a six-year-old's party? Bill looked horrified. It adds dramatic flair. Now, let us discuss the getaway vehicle. Please tell me it is not the ice cream truck again. It is the perfect camouflage! Who suspects an ice cream truck at a party? Everyone! Bill yelled. It is the only thing they want! Exactly. We blend in. Now, do you have the burlap sacks? I have two laundry bags. Will that work? Acceptable, Molly conceded. But if there is a giant teddy bear, we might need a dolly. I cannot believe I am doing this. Stop whining and start practicing your innocent party guest face, Molly commanded. Bill made a face that looked more like he had swallowed a lemon. Terrible, she sighed. Just blend in and don't touch the cake. Wait, we can't eat the cake? Focus, Bill! The mission comes first. Wait, what about the dog? Bill asked. The neighbor mentioned a Doberman. A mere distraction, Molly dismissed. I have a bag of premium organic beef jerky. He will be too busy negotiating his soul for a snack to notice us. I still think this is a low point for our careers. It is a strategic pivot, Bill. Now, put on the oversized polo shirt. We leave in five.
created_at
"2026-05-10T13:26:26.521Z"
5/10/2026, 1:26:26 PM
model
"google/gemma-4-31b-it-20260402"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
215
moderation_latency
(null)
generation_time
68327
tokens_prompt
77
tokens_completion
2472
native_tokens_prompt
82
native_tokens_completion
2430
native_tokens_completion_images
(null)
native_tokens_reasoning
1787
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"
service_tier
(null)
usage
0.00093406
router
(null)
provider_responses
0
endpoint_id
"4149ab03-3b69-45a0-b94c-6c253b8641a1"
id
"chatcmpl-RjqS7XGq7jOdaDMTZ5C1Ui4Z"
is_byok
false
latency
97
model_permaslug
"google/gemma-4-31b-it-20260402"
provider_name
"DeepInfra"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1778419586-GE3VBtxF2VAtwAk1Repn"
session_id
(null)
api_type
"completions"
id
"gen-1778419586-AAQIxgi5A4FDxDFj6opU"
upstream_id
"chatcmpl-RjqS7XGq7jOdaDMTZ5C1Ui4Z"
total_cost
0.00093406
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
441
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 70.00%
neededClean
false
wordsTotal
450
wordsDialogue
0
0.0000%