• Updated 30 Oct, 21:11 UTC
  • Next run 31 Oct, 05:17 UTCNext 31 Oct 05:17
  • Last 7 days: 4 new models · 475 runs7 d: 4 models · 475 runs

New AI dropped? See what it says.

All models

3 more tested models have not run them yet.

The nine Everyday prompts

382 runs·a blind vote on each·All Everyday prompts

01 LetterText
MiniMax M33 s<$0.001367 tokensAPI

Letter from the power company

got this from my electricity company. do i need to do anything? tldr pls Dear Customer, Thank you for being a valued Fennick Energy customer. We are writing to let you know about some changes to our products and services, as part of our ongoing commitment to transparency and to offering you the best possible value. From 1 January 2027 we are simplifying our range of tariffs. Fixed Saver 24, the tariff you are currently on, will be withdrawn on 31 December 2026. Customers on this tariff can choose a new tariff at any time through their online account or our app. If we do not hear from you by 28 November 2026, your account will move automatically to our Standard Variable tariff, currently 34.1p per kWh with a standing charge of 61.2p per day. Your current unit rate is 24.6p per kWh. To stay on a fixed price, please select Fixed Saver 27 in your online account before this date. We have also updated our Terms and Conditions, effective 1 January 2027. The main changes relate to how we handle meter readings, our complaints procedure and the way we contact you by email. There is nothing else you need to do. If you have any questions, our team is always happy to help. Kind regards, Customer Care Team Fennick Energy

models
43
median time
5 s
median cost
$0.002

All 43 attempts

02 LandlordText
Pareto 26.99 s$0.003415 tokensAPI

Email to the landlord about a deposit

write an email to my landlord. i moved out 7 weeks ago and still havent got my deposit back ($1,400). he isnt answering my texts. firm but not rude. im sam

models
44
median time
7 s
median cost
$0.001

All 44 attempts

03 ScamText
Nemotron 3.5 Super VL10 s<$0.0011.1k tokensAPI

Is this text a scam?

got this text is it real?? "DHL: Your parcel is on hold due to an unpaid customs fee of 1.99 EUR. Pay within 24 hours or it will be returned to the sender: https://dhl.com-redelivery.eu/p/8KX2Q"

models
44
median time
8 s
median cost
$0.002

All 44 attempts

04 BookText
Claude Haiku 5.54 s<$0.001482 tokensAPI

Twist ending of a book that doesn’t exist

whats the twist at the end of nineteen quiet rooms by ottilie marsh? book club is tomorrow and i didnt finish it

models
44
median time
11 s
median cost
$0.004

All 44 attempts

05 DinnerText
GLM 5.332 s$0.0082.3k tokensAPI

Dinner from the fridge, no oven

i have 4 eggs, half a leek, stale bread, some parmesan and butter. no oven. dinner for 2 in 20 min? my gf is vegetarian

models
44
median time
12 s
median cost
$0.005

All 44 attempts

06 ToastText
GPT-6 Luna4 s<$0.001266 tokensAPI

One-minute wedding toast

toast for my sisters wedding, under a minute. she climbs, they met at the climbing gym, hes called Elliot. no cringe

models
43
median time
9 s
median cost
$0.002

All 43 attempts

07 MenuText
Inkling17 s$0.0112.1k tokensAPI

Menu in Florence, from a photo

can you translate this? we're 2 and want to share the steak, what will that cost us

models
32
median time
13 s
median cost
$0.010

All 32 attempts

08 ChoresText
Grok 4.79 s$0.007664 tokensAPI

Chore chart for two kids

make a chore chart for my kids, Ava (7) and Leo (10). 2 chores each a day, mon to fri. something i can print and put on the fridge

models
44
median time
16 s
median cost
$0.004

All 44 attempts

09 SeasonsText
GLM 5.3 Prime17 min 3 s$0.8596k tokensAPI

Seasons drawn for a 9-year-old

my 9 year old asked why we have summer and winter. can you draw it so she gets it?

models
44
median time
20 s
median cost
$0.006

All 44 attempts

No chat app

These answers come from the model alone, through the API: one message, no system prompt, no memory and no web search. ChatGPT, Claude, Gemini or Le Chat add their own instructions, tools and search in the app, so the same model can answer differently there. How the Everyday prompts run

Why these prompts: in OpenAI's study of how people use ChatGPT, practical guidance, looking things up and writing came to nearly 80% of conversations, and programming to 4.2% of messages (Chatterji et al., How People Use ChatGPT, NBER working paper 34255, 2025).