NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-01-01M1EMFHFPNSVNW30B3XH5YZQ3
Run Details
Model:
Thinking Machines Inkling Small (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 1s
Cost:
$0.003983
Provider:
Together
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.30%
Dialogue to Total Word Ratio
(details)
0.1486%
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 tapped the invitation on the dashboard. "Bill, I cannot believe you talked me into breaking into a children's party to steal presents." Bill adjusted his camouflage beret. "Toddler presents are high-value, Molly. Those loot bags have iPads woven into the plastic. They're basically party favors for parents who have given up on reality." "These are six-year-olds, Bill. They have emotional intelligence now." "They have emotional stuff to trade for cake. We're not hurting anyone. We're redistributing wealth to people who actually appreciate a good heist." Molly sighed. "The invitation says 'no uninvited ninjas.'" "That's not a rule. That's a theme." Bill pulled out a crumpled map. "Plan is simple. We enter through the balloon arch at nine-thirty. The cake is a diversion. While they're singing, we hit the gift table." "The gift table is in the same room as the cake, Bill. There's no hallway." "We use the bouncy castle as our vertical insertion point." "You mean jumping?" "I mean tactical elevation." Molly shook her head. "What about the clown? Clara said he has security training." "Clowns are a psychological operation. We deploy silent gas." "That's banned by the Geneva Convention." "Not in clown jurisdiction." Bill pointed at a photo of the venue. "See the backyard fence? We cut through the hydrangeas. Low visibility, high shrubbery." "The hydrangeas are specifically mentioned in the neighborhood newsletter." "Then we go through the garage. The garage code is the birthday boy's last name plus one-two-three." "What's his last name?" "Kid's name is Oliver. But the garage says 'Olaf.'" "Olaf?" "Maybe he's Norwegian." Molly pulled out her phone. "I just texted my mom asking if she knows a six-year-old Olaf who lives on Maple." "She's responding in twelve minutes." "We don't have twelve minutes. The party starts at two, and we need to move during the piñata phase because everyone is screaming." Bill nodded. "The piñata is our audio cover. When they break it open, they won't hear the duct tape." Molly stared at him. "You bought duct tape?" "It's professional grade." "It says 'colorful' on the roll." "That makes it festive." Bill grabbed a sack. "Once inside, we grab by weight, not sentiment. Those stuffed animals are heavy. The plastic trophies are lightweight but high resale." "Molly, this is immoral." "Correct. That's why I'm only taking one thing." "Which is?" "The big present from the grandparents. The one with the bow." "Why?" "It's probably a tablet, and tablets are easier to fence than tears." Bill smirked. "You're getting soft." "I'm getting practical. If we steal everything, we cause a trauma. If we steal the one big thing, the party continues, and we have a new laptop." "That doesn't make sense. A new laptop isn't a present." "We sell it." "We sell a stolen laptop to buy more duct tape?" "Exactly." Bill pulled out a hairband. "Fine. But we need disguises." "We're two elderly volunteers?" "We already have the vests." Molly looked at her vest, which said "Party Helper" in glitter. "This vest is suspicious." "Clowns are suspicious. Volunteers are invisible." "We're not elderly." "You're thirty-five and I have a mustache. That's elderly in party years." Molly sighed. "Okay. We go in. We blend. We hit the table when they do the icebreaker." "What's the icebreaker?" "It's called 'Pass the Parcel.'" "Pass the parcel is not an icebreaker. It's a torture device." "Exactly. Everyone is distracted." Bill grabbed a present-size box from the trunk. "I'm putting this on as a prop. If anyone asks, it's a gift." "Bill, that's a gift box." "It's empty inside. I'm a method actor." Molly opened the door. "Let's go. And if I see a clown, I'm leaving." "You're leaving because you don't like the clown?" "I'm leaving because you brought glitter glue." "That's for the getaway car." "You put glitter glue on the car?" "It matches the balloons." Molly shook her head. "You're insane." "I'm professional." "You're both." They walked toward the party. Bill looked back at the house. "Wait. Did we check if there's a dog?" "Only if it's a party dog." "Party dogs guard presents." "We treat it with the leftover cake." Molly knocked. "If anyone asks, we are here to deliver the balloon tower." Bill held up a stack of balloons. "This is my tower." "That's a bunch of balloons tied to your head." "I've been practicing." They entered. Inside, children screamed in delight or terror. Bill whispered. "I see the table." Molly whispered back. "I see a parent with eyes like a hawk." "That's Mrs. Henderson. She organizes the neighborhood watch." "She's also holding a camera." "We don't make eye contact with cameras." "We do if we're disguised." "We're not." "Then we improvise." Bill grabbed a cupcake. "I'm eating this for cover." "Bill, that's for a child named Oliver." "He can have the footage." Molly slid past them toward the presents, clutching the empty box. Bill stopped her. "Wait. Do we have a plan for the exit?" "We're going to run?" "No, we're going to blend back in and exit with the balloon tower." "That makes no sense, but it's the only plan we have." Molly reached for the biggest present with gold ribbon. Bill grabbed her arm. "Molly, that's not a present." "It's a present." "It's a decoy. The parents put it there to catch thieves." "Then it's a gift for us." "No, it's a baby shower gift with a pacifier inside." "Still useful?" "No." Molly looked at Bill. "We should leave." Bill nodded. "We should leave. And if anyone asks, we were never here." "Good." "And if they ask about the presents?" "We say we brought them." Molly grabbed the gold ribbon present anyway. Bill grabbed a bag of loot. They walked out, past the balloon arch, into the hydrangeas. Bill turned to Molly. "We have one present and a bag of party favors." Molly looked inside the bag. "And a juice box." "We can sell the juice box for five dollars." "We are heroes of crime." "Only if no one catches us." "Molly?" "Bill?" "We forgot the duct tape." "We used the glitter glue instead." "Good." They walked away, leaving the party slightly less complete, slightly more glittered, and completely confused.
created_at
"2026-09-01T14:02:59.968Z"
9/1/2026, 2:02:59 PM
model
"thinkingmachines/inkling-small-20260730"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2336
moderation_latency
(null)
generation_time
60529
tokens_prompt
75
tokens_completion
3794
native_tokens_prompt
75
native_tokens_completion
3288
native_tokens_completion_images
(null)
native_tokens_reasoning
1866
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
0.0039831
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788271379-tujanuM3aHsXXuv4fQT8"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788271379-rj8rJmsrCzaGppQWZoJF"
upstream_id
"2aef0a35-aws_ue1"
provider_responses
0
endpoint_id
"0ed263b7-f746-4eb6-a14c-54f2e1cc2269"
id
"2aef0a35-aws_ue1"
is_byok
false
latency
2331
model_permaslug
"thinkingmachines/inkling-small-20260730"
provider_name
"Together"
status
200
total_cost
0.0039831
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Together"
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
1031
0.30%
Dialogue to Total Word Ratio
Ratio: 80.79%, Deviation: 10.79%
neededClean
false
wordsTotal
1041
wordsDialogue
841
0.1486%