NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-01-01M1EMZMNRGV4K346X6M9DKRQB
Run Details
Model:
Thinking Machines Inkling Small (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
40.4s
Cost:
$0.006864
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.01%
Dialogue to Total Word Ratio
(details)
0.0058%
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
Molly leaned against the lamppost and adjusted her fedora. "Bill, are you telling me the target is a birthday party?" Bill nodded, clutching a crumpled invitation. "The Henderson kid. Six years old. Gifts worth forty bucks in generic plastic dinosaurs." "We're not taking dinosaurs," Molly said. "We're taking the moment." Bill scratched his chin. "The moment is a piƱata full of candy. The presents are stacked in the garage. I checked. They have a slide projector showing Finding Nemo on loop." "Okay," Molly said. "We go in through the bounce house. It's inflatable, therefore unpickable. We cut a hole with a kitchen knife." "That's violent," Bill said. "It's a birthday party, Bill. Violence is the theme." They walked around the block. "What about the clown?" Bill asked. Molly sighed. "The clown is on our payroll. He sees us, he forgets his nose. It's basic psychology." "Basic fraud," Bill corrected. "Same difference," Molly said. "We need disguises. I say we dress as grandparents. Grey hair, orthopedic shoes, and the energy of someone who has lived through three wars." "Grandparents don't squeeze through bounce house seams," Bill pointed out. "We don't squeeze," Molly said. "We detonate. Small charge. No one notices over the screaming." Bill pulled out a notepad. "The timeline: arrive at 4:30, present opening at 4:45, cake at 5:00, we leave with the loot by 5:15, before the parents realize the good presents are gone." "Why would parents realize?" Molly asked. "Because kids cry," Bill said. "Kids always cry," Molly said. "That's not evidence, that's Tuesday." They stopped at a dumpster behind the venue. "The security is mother-in-chief Janet," Bill whispered. Molly grinned. "I know Janet. She owes me five hundred dollars from a poker game in 2014." "She also has a taser in her purse," Bill said. "Then we'll use her purse as the target," Molly said. "Steal the taser, steal the presents, leave a note that says Happy Birthday, Love, The Grandparents." Bill looked at her. "That's genius." "It's also criminal," Molly said. "Same difference," Bill said. They entered through the backyard. The bounce house hummed like a dying whale. Molly knelt, knife in hand. "Wait," Bill said. "Is that the clown?" Molly looked up. A man in oversized shoes was juggling flaming batons near the sandbox. "That's not our clown," Molly said. "That's a clown with ambition." "Should we abort?" Bill whispered. "Too late," Molly said. "We're already inside the perimeter. The parents think we're the grandparents who brought the dinosaur-shaped cake." Bill nodded. "Right. The cake." "Exactly," Molly said. "We don't steal the cake. The cake is a distraction. We steal the presents while they sing Happy Birthday." They reached the garage. Presents were stacked like a gift-wrapped tower. "How many?" Bill asked. "Twenty-seven," Molly said. "Mostly Nerf guns and one science kit that probably requires adult supervision." "Adult supervision is what we're avoiding," Bill said. Molly cut open a sack. "I see a box for a remote-controlled robot. That's the prize. The rest is filler." Bill loaded his sack. "What if they notice?" "We leave a replacement," Molly said. "We put in socks. Actual socks. The kind with dinosaurs on them. They will never know." They ran back to the bounce house. The clown was now doing a magic trick with a rabbit that had clearly had enough. "Let's go," Molly said. Bill tossed a present over his shoulder. "What about Janet?" Molly glanced at the mother-in-chief, who was arguing with a magician about balloon animals. "She owes me money. She won't tinker with the bounce house." They slipped out through the hedge, into the alley. The presents clinked. Bill turned to Molly. "We did it." "We did," Molly said. "But next time, let's pick a party with better snacks." "Next time, let's pick a party that's not a birthday," Bill said. "Next time," Molly said, "we're professionals." They disappeared into the night, leaving behind only the faint smell of cake, regret, and an inflatable bouncy castle that would never be the same.
created_at
"2026-09-01T14:11:47.525Z"
9/1/2026, 2:11:47 PM
model
"thinkingmachines/inkling-small-20260730"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
209
moderation_latency
(null)
generation_time
40383
tokens_prompt
75
tokens_completion
6031
native_tokens_prompt
74
native_tokens_completion
5692
native_tokens_completion_images
(null)
native_tokens_reasoning
4978
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
0.0068637
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788271907-YvRZPnxeW4Jg21jtQsSv"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788271907-GvgYZA7P4xMhEkdyx27u"
upstream_id
"chatcmpl-RCqMfutgXBKhzvyM3KBkYjha"
provider_responses
0
endpoint_id
"d1d9657b-3c3a-49b5-8f81-33204e7c71c3"
id
"chatcmpl-RCqMfutgXBKhzvyM3KBkYjha"
is_byok
false
latency
91
model_permaslug
"thinkingmachines/inkling-small-20260730"
provider_name
"DeepInfra"
status
200
total_cost
0.0068637
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
668
0.01%
Dialogue to Total Word Ratio
Ratio: 63.46%, Deviation: 13.46%
neededClean
false
wordsTotal
676
wordsDialogue
429
0.0058%