NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-10-01KR8VH5EDC40TN5V8VP4GM9WA
Run Details
Model:
Gemma 4 31B
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
18.7s
Cost:
$0.000234
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 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
Bill spread a crayon-drawn map across the kitchen table. He pointed a greasy finger at a circle labeled The Danger Zone. "The bounce house is the primary obstacle, Molly," Bill whispered. Molly stared at him. "It is a inflatable castle, Bill. Not a fortress." "That is exactly what the manufacturers want you to think," Bill countered. "One wrong step and you are tossed into a sea of toddlers. It is chaos in there." Molly sighed and leaned over the map. "Focus. We enter through the side gate at 14:00 hours. I will create a diversion with the cake." "The cake?" Bill asked. "That seems risky." "I will accidentally trip and send the buttercream flying," she explained. "While the parents are scrubbing frosting off the patio furniture, you swoop in." "Swoop in where?" "The gift table, you idiot. It is positioned right next to the juice boxes." Bill frowned. "What if there are guards?" "Guards? It is a six-year-old's birthday party." "Kids are the best guards," Bill insisted. "They are small, fast, and they scream at a frequency that can shatter glass. They will spot us instantly." Molly rubbed her temples. "You are wearing a giant panda suit, Bill. You will blend right in with the entertainment." "I thought the panda suit was too obvious," Bill admitted. "I was considering a very large balloon animal costume." "The panda is fine. Just get in, bag the Lego sets, and get out." "What about the stuffed animals?" Bill asked. "Do we take the plushies or leave them?" "Leave the plushies. They are bulky and low-value," Molly snapped. "We are after the high-end electronics and the gift cards." "I just feel like leaving the plushies is cold," Bill murmured. Molly glared at him. "Are you a professional thief or a nursery school teacher?" "I can be both," he replied proudly. "Just promise me you won't get distracted by the snack table," she warned. Bill looked thoughtful. "Will there be pigs-in-a-blanket?" "Yes, Bill." "Then I cannot make any promises regarding my professional integrity." Molly groaned and slumped into her chair. "I cannot believe I am partnering with a man who is intimidated by a bounce house." "It is the physics of the bounce, Molly! It is unpredictable!" "Just put on the suit," she sighed. Bill stood up with a look of grim determination. "For the Legos." "For the Legos," she whispered, wondering if it was too late to join a circus.
created_at
"2026-05-10T11:49:30.197Z"
5/10/2026, 11:49:30 AM
model
"google/gemma-4-31b-it-20260402"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1035
moderation_latency
(null)
generation_time
18665
tokens_prompt
77
tokens_completion
643
native_tokens_prompt
79
native_tokens_completion
590
native_tokens_completion_images
(null)
native_tokens_reasoning
0
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
0.00023447
router
(null)
provider_responses
0
endpoint_id
"4149ab03-3b69-45a0-b94c-6c253b8641a1"
id
"chatcmpl-RjczRfUe55KxVkHNMG1L6wDN"
is_byok
false
latency
50
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-1778413770-UU1CDahDcIl7FrnvX9rO"
session_id
(null)
api_type
"completions"
id
"gen-1778413770-nWsusK5OgpDWK1i9DZG9"
upstream_id
"chatcmpl-RjczRfUe55KxVkHNMG1L6wDN"
total_cost
0.00023447
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
406
0.00%
Dialogue to Total Word Ratio
Ratio: 76.09%, Deviation: 26.09%
neededClean
false
wordsTotal
414
wordsDialogue
315
0.0000%