NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-10-01KR8W7KYHEMMQAFVJNJ2AWJKN
Run Details
Model:
Gemma 4 26B
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
13.6s
Cost:
$0.000316
Provider:
Io Net
Evaluation summary
Score
Evaluator
Details Link
0.49%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.2460%
Overall
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 blueprints across the sticky laminate table of the diner, her eyes narrowed with the intensity of a general planning an invasion. Bill sat opposite her, struggling to pierce a singular, lonely blueberry with a plastic fork. The target is the Miller residence, Molly whispered, her voice low and gravelly. Saturday, two p.m. High density of colorful wrapping paper. High probability of loot. Bill swallowed the blueberry with a grimace. Is there a perimeter? Are we talking laser grids or just toddlers in dinosaur costumes? Molly tapped a finger on a red circle near the backyard patio. It is a tactical nightmare, Bill. The terrain is cluttered with inflatable bouncy castles and a slip-and-slide that acts as a natural lubricant for security evasion. If you slip, the noise will be deafening. I can handle a little water, Bill said, leaning in. What about the primary vault? The gift table, Molly corrected sharply. It sits centrally in the living room, flanked by a mountain of juice boxes and a bowl of lukewarm potato chips. We have a four-minute window between the cake cutting and the inevitable sugar-induced meltdown. Four minutes? Bill scoffed, waving his fork. That is barely enough time to scout the terrain. What if a parent catches us? Molly sighed, rubbing her temples. We have the disguises. You are the eccentric Uncle Barney, and I am the overly enthusiastic clown instructor. If anyone asks questions, we pivot immediately to organized singing. Uncle Barney? Bill muttered. I don't do polka, Molly. It’s bad for my street cred. Just wear the Hawaiian shirt and keep your mouth shut, she snapped. We move in via the side gate. I will create a diversion by accidentally tripping a toddler or spilling a bowl of fruit punch. While the chaos unfolds, you sweep the presents into the tactical duffel bags. Do we have a way to transport the goods? Bill asked, his eyes brightening. I am thinking a large, empty wagon. One of those heavy-duty ones that can handle the weight of several dozen Lego sets. A red wagon with a squeaky wheel, Molly said, checking a box on her notepad. It is perfect. It blends into the aesthetic of a suburban backyard. We will look like part of the decor. Bill nodded slowly, the gravity of the mission finally sinking in. And if we hit the jackpot? If there is a high-end gaming console in there? We split it fifty-fifty, Molly said, her gaze hardening. But if you get distracted by the chocolate fountain again, I am leaving you to face the wrath of a dozen five-year-olds alone. I make no promises about the fountain, Bill admitted. Molly stood up, gathering her maps. Then sharpen your senses, Bill. This is the big one. The heist of the century, Bill whispered, finally conquering the blueberry.
created_at
"2026-05-10T12:01:45.945Z"
5/10/2026, 12:01:45 PM
model
"google/gemma-4-26b-a4b-it-20260403"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
314
moderation_latency
(null)
generation_time
13544
tokens_prompt
77
tokens_completion
719
native_tokens_prompt
79
native_tokens_completion
609
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
0.00031635
router
(null)
provider_responses
0
endpoint_id
"0bc24d30-bd26-4fa8-968b-260c0011a4a5"
id
"chatcmpl-94ab882733b275a7"
is_byok
false
latency
314
model_permaslug
"google/gemma-4-26b-a4b-it-20260403"
provider_name
"Io Net"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1778414505-OMDsb4JgNCDHsG2Ft96F"
session_id
(null)
api_type
"completions"
id
"gen-1778414505-EGoZgHmRldtm9bo4ZUdM"
upstream_id
"chatcmpl-94ab882733b275a7"
total_cost
0.00031635
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Io Net"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.49%
Matches word count
n/a
neededClean
false
words
473
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 30.00%
neededClean
false
wordsTotal
482
wordsDialogue
0
0.2460%