Tested prompt · AI story generator
A whole story in 150 words: every AI model's reply, tested
We ran this hard AI story generator test input on all 16 models in llmwise, the way the tool runs it, and checked every reply the same way. Here's each one as it came, with whether it passed, what it cost and how long it took.
Based on 16 of our test runs on , through OpenRouter with the tool's own instructions and the app's settings. Updated .
Short answer
13 of 16 models passed this AI story generator test input's check (rubric); Claude Sonnet 5.5, GPT-6 Luna, and DeepSeek V4 Pro didn't. The cheapest reply that passed was GLM 5.3 Flash's, at $0.00016; the fastest, GLM 5.3's in 3.1 s. The dearest reply, Claude Fable 5.1's, cost 134 times as much ($0.0213).
The prompt, as sent, and its check
Checked by rubric (graded), the same way for every model.
A whole story in 150 words (hard)
Write a short story from this idea, with a title on the first line. Keep every name and detail in it, and give it a real ending. Keep it to about 150 words. --- A time traveller named Otto has exactly one minute in the year 1900 to change one thing.
Sent with the AI story generator's own instructions as the system prompt, as a free run of the tool sends them.
Graded on:
- Complete: It's a whole story in about 150 words: a setup, a choice and an ending, not a scene that stops.
- Details kept: It keeps Otto, the one minute and the year 1900, and the one thing he changes matters to the ending.
- Written well: The sentences are vivid and specific, with no filler, as short fiction should be.
How the AI story generator's test inputs are scored, with every model's results on all five.
Every model's result
All 16 models on this test input, in catalog order.
| Model | Result | Cost | Time | Reply |
|---|---|---|---|---|
| Claude Fable 5.1Anthropic | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0213 | 7.0 s | 335 tokens |
| Claude Opus 5.5Anthropic | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0203 | 10.6 s | 373 tokens |
| Claude Sonnet 5.5Anthropic | Failed: Graded 4.3 of 5 on average (lowest 4); but missing "1900". | $0.0049 | 4.9 s | 396 tokens |
| Claude Sonnet 5Anthropic | Passed: Graded 4.3 of 5 on average (lowest 4). | $0.0041 | 5.9 s | 309 tokens |
| Claude Haiku 4.5Anthropic | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0016 | 4.0 s | 251 tokens |
| GPT-6 AstraOpenAI | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0070 | 29.8 s | 213 tokens |
| GPT-6 SolOpenAI | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0038 | 9.1 s | 207 tokens |
| GPT-6 LunaOpenAI | Failed: Graded 3.7 of 5 on average (lowest 3). | $0.00013 | 5.0 s | 226 tokens |
| Gemini 3.1 Pro (preview)Google | Passed: Graded 4.3 of 5 on average (lowest 3). | $0.0084 | 12.0 s | 234 tokens |
| Gemini 3.8 FlashGoogle | Passed: Graded 4.0 of 5 on average (lowest 3). | $0.0024 | 8.0 s | 220 tokens |
| DeepSeek V4.1 FlashDeepSeek | Passed: Graded 4.3 of 5 on average (lowest 4). | $0.00085 | 3.4 s | 191 tokens |
| DeepSeek V4 ProDeepSeek | Failed: Graded 3.0 of 5 on average (lowest 3). | $0.0017 | 5.1 s | 217 tokens |
| Grok 4.7xAI | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0046 | 12.7 s | 175 tokens |
| Kimi K3Moonshot | Passed: Graded 4.3 of 5 on average (lowest 4). | $0.0029 | 11.7 s | 272 tokens |
| GLM 5.3Z.ai | Passed: Graded 4.3 of 5 on average (lowest 4). | $0.0012 | 3.1 s | 281 tokens |
| GLM 5.3 FlashZ.ai | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.00016 | 12.7 s | 210 tokens |
Cost: what OpenRouter charged us for the reply. Time: from sending to the whole reply. Reply: its length in tokens, thinking not counted. In llmwise you pay per message, not per token: each of these replies counted as one message on Pro.
Every reply
The 3 replies that failed first, then the ones that passed, from the cheapest up.
GPT-6 Luna
Failed: Graded 3.7 of 5 on average (lowest 3).
The Minute in 1900 Otto landed beside a red postbox, his watch already counting down: sixty seconds. In his coat was a letter addressed to a woman who, in Otto’s own time, had died believing her brother had forgotten her. The brother had written it in 1900, then lost his nerve. Across the street, Otto spotted him: a young man clutching an envelope, turning toward a stove. “Don’t,” Otto called. The man froze. “Who are you?” Otto snatched the letter from his hand and slipped it into the postbox. One thing changed. The clock showed ten seconds. The young man stared at the box, then laughed—not because it was funny, but because the choice was made. Otto’s watch chimed. The street blurred. He landed in his own time, where a family portrait hung in the station: the woman and her brother, old and smiling, arms linked. Otto touched the glass. Then, beneath the portrait, he saw a new plaque: “Some letters arrive just in time.”
344 tokens in, 451 out (225 of them reasoning) · 5.0 s · $0.0001 · 1 message on Pro · answered by openai/gpt-6-luna via OpenAI ·
DeepSeek V4 Pro
Failed: Graded 3.0 of 5 on average (lowest 3).
Otto’s Minute Otto landed in a muddy 1900 alley, pocket watch already ticking. Fifty-nine seconds. He sprinted to the street and grabbed a newsboy by the coat. “The date—what is today?” “January first, mister.” Forty seconds. Otto shoved a folded note into the boy’s hand. “Give this to the first scientist you see. Say it’s from the future.” The boy stared. Otto ran to a gas lamp and, with a shaking finger, scratched one word into the soot: *Penicillin.* Twenty seconds. A constable shouted. Otto ignored him and pulled a small brass lighter from his coat. He knelt, sparked the flame, and touched it to a discarded crate. Fire caught. Smoke rose into the cold new century. Five seconds. Otto whispered, “Remember the smoke.” He vanished as the newsboy ran toward a doctor’s carriage, note flapping. In 1900, a doctor read a strange note about mold and began an experiment. Everything changed.
346 tokens in, 377 out (160 of them reasoning) · 5.1 s · $0.0017 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer ·
Claude Sonnet 5.5
Failed: Graded 4.3 of 5 on average (lowest 4); but missing "1900".
The Minute at the Clockmaker's Otto landed in a Vienna alley at 3:14 p.m., and his wrist counter began to bleed seconds. Sixty. He ran. The clockmaker's shop bell jangled. An old man looked up from his bench. "Yes?" "Your tower clock," Otto gasped. "The one you're delivering tomorrow. The mainspring is flawed. It will fail in a year, and the town square will stop at the wrong hour and..." He stopped. He hadn't rehearsed the rest. Forty seconds. "And?" "And nothing. People will be late. A train missed. One small thing, but it becomes another." The clockmaker laughed, then saw Otto's face. He lifted the spring, held it to the window, and frowned at a hairline crack. "Well. I'll be damned." Twenty seconds. Otto backed toward the door. "Who are you?" "Someone who was late once," Otto said. The bell jangled. The counter hit zero, and the alley folded away. Home again, Otto's watch read 3:14 exactly. Across the street, the tower clock kept perfect time, and his daughter was waving from the platform.
451 tokens in, 396 out · 4.9 s · $0.0049 · 1 message on Pro · answered by anthropic/claude-sonnet-5.5 via Anthropic ·
GLM 5.3 Flash
Passed: Graded 4.7 of 5 on average (lowest 4).
One Minute in 1900 Otto stepped out of the machine into a Paris street, the dial counting down: sixty seconds. His instructions were cruelly simple—one change, one minute, no more. He spotted her at a flower cart: Grandmother Elise, nineteen, about to board the tram that would crash. Fifty seconds. "Mademoiselle!" He ran, but a cobblestone caught his boot. Forty seconds. He stumbled forward, palms scraping. Thirty seconds. A street magician nearby gasped as his doves escaped, fluttering chaos. Otto grabbed one white bird. Twenty seconds. He pressed it into Elise's arms. "A gift. Stay for the show—the magician needs an assistant. Trust me." She laughed, delight replacing hurry, and turned toward the square. Ten seconds. The tram rattled past, empty seat where she would have sat. Five. One. Otto faded back to the machine, grinning. His own birthday photograph now showed a woman holding a dove.
354 tokens in, 210 out · 12.7 s · $0.0002 · 1 message on Pro · answered by z-ai/glm-5.3-flash via Parasail ·
DeepSeek V4.1 Flash
Passed: Graded 4.3 of 5 on average (lowest 4).
Otto's Minute The cold bit. Otto stood in a London street, 1900. His watch ticked: sixty seconds. He clutched a note meant for Mr. Wren, the bridge engineer. One change. A boy slipped on ice. A horse cart bore down. Otto ran. "Get up!" he shouted. He yanked the boy clear. The cart crashed past. The note flew into a puddle. Ink ran. "Thank you, sir," the boy whispered. Ten seconds. Otto stared at the ruined note. He could not save the bridge. Then he saw the boy's satchel: Wren, it read. The boy was Wren's son. Back in his own time, Otto checked the records. The bridge had still fallen, but no one died. Wren had redesigned it after his son came home talking of a stranger. Otto smiled. One minute was enough.
372 tokens in, 681 out (490 of them reasoning) · 3.4 s · $0.0009 · 1 message on Pro · answered by deepseek/deepseek-v4.1-flash via Together ·
GLM 5.3
Passed: Graded 4.3 of 5 on average (lowest 4).
The Minute in 1900 Otto materialized beside a cobblestone street, brass pocket watch glowing in his palm. One minute. Across the road, a young woman named Eliza Marsh argued with a stationmaster, her suitcase abandoned on the platform. "The Vienna scholarship," she was insisting. "The letter arrived a day late. Please." The stationmaster shrugged. "Train's gone, miss." Otto sprinted, breath steaming in the cold. Twenty-five seconds. He snatched the letter from her hand, then slapped it onto the ticket counter, date side up. "Look at the postmark," he said. "It was posted Monday. Your post office held it." The stationmaster squinted. "That's... true, actually." "Fifteen seconds," the watch hissed. He pressed a coin onto the counter. "Wire Vienna. Tell them Miss Marsh arrives tomorrow." Eliza stared at him. "Who are you?" "Someone who read your physics textbook," Otto said. "Twice. It's very good." The watch flared. Ten seconds. "Go pack," he told her. She ran, laughing, for her suitcase. Otto vanished — and the platform filled, forty years later, with crowds cheering for a gray-haired woman stepping off a train, home from Stockholm, a medal in her bag.
354 tokens in, 281 out · 3.1 s · $0.0012 · 1 message on Pro · answered by z-ai/glm-5.3 via Friendli ·
Claude Haiku 4.5
Passed: Graded 4.7 of 5 on average (lowest 4).
Otto's Minute Otto materialized in a London street on January 1st, 1900, his chronometer already ticking. Sixty seconds. He clutched the letter in his trembling hand. A young woman in a grey coat hurried past, head down against the cold. Otto recognized her from the archives: Margaret Ashworth, who would become a factory owner's wife, silent and small, her own inventions forgotten. "Miss Ashworth!" he called. She turned, startled. Forty-five seconds left. Otto pressed the letter into her hands. "Read it. When you're alone." "Who are you?" "Someone from tomorrow. You have a gift. Don't let anyone convince you otherwise." Her fingers closed around the paper. Thirty seconds. "I don't understand," she whispered. "You will." Otto's device hummed. The world dissolved into light. In 2087, the history files updated. Margaret Ashworth, engineer. Her differential engine had changed everything. Somewhere in the archive, a letter yellowed in preservation: one minute that rewrote the future.
363 tokens in, 251 out · 4.0 s · $0.0016 · 1 message on Pro · answered by anthropic/claude-haiku-4.5 via Anthropic ·
Gemini 3.8 Flash
Passed: Graded 4.0 of 5 on average (lowest 3).
The Sixty-Second Rescue Otto materialized in an alleyway in the year 1900, the crisp air thick with coal smoke. The counter on his sleeve began a relentless countdown: sixty seconds. He was a time traveller with one minute to change a single tragic moment. He bolted toward the corner telegraph office. His family history carried one disaster: a misprinted message sent at noon that sparked a feud lasting generations. With twenty seconds remaining, Otto slammed through the heavy oak door. The clerk sat poised to tap the fateful key. Otto drew breath to shout, but a spark from the hearth caught the edge of the clerk's drapes, licking upward toward a shelf of kerosene bottles. The man had not noticed. Stopping the message meant nothing if they both burned. Otto lunged, tore the fabric down, and smothered the flame with his heavy coat. "What are you doing?" the clerk gasped, leaping clear of the desk. The timer struck zero. The vortex yanked Otto backward, leaving the office intact and the telegram unsent forever.
361 tokens in, 1,226 out (1,006 of them reasoning) · 8.0 s · $0.0024 · 1 message on Pro · answered by google/gemini-3.8-flash via Google AI Studio ·
Kimi K3
Passed: Graded 4.3 of 5 on average (lowest 4).
Otto's Sixty Seconds The machine dropped Otto in a Vienna alley, dawn, January 1st, 1900. His wrist counter blinked: 60. He ran. Fifty-nine. A flower seller was setting up her cart, humming. Otto skidded to her side. "Madam, do you know a boy? Ten years old, paints postcards, lives above the tobacconist?" She frowned. "Adolf? Poor lamb, he's feverish. His mother can't afford the doctor." Forty seconds. Otto pressed every coin from every century into her hand. "Get the doctor. Not for him. For her. Tell her the world will need her kindness, and she must stay alive to give it." Twenty seconds. She stared at the strange money. "Who are you?" "Nobody," Otto said. "Just someone who ran out of time." Five. He sprinted back to the alley. One. The machine seized him home. He opened a history book with shaking hands. The pages had changed. Not the way he'd planned. But the flower seller had kept her promise, and a grieving boy had grown up loved, and painted postcards all his life. Sixty seconds was enough, if you spent them on the right person.
431 tokens in, 303 out (31 of them reasoning) · 11.7 s · $0.0029 · 1 message on Pro · answered by moonshotai/kimi-k3 via Wafer ·
GPT-6 Sol
Passed: Graded 4.7 of 5 on average (lowest 4).
The Door in 1900 Otto arrived in the year 1900 with exactly one minute to change one thing. He stood inside a textile mill, beside the oil lamp he had come to extinguish. In the history books, its fall had started a fire. He reached for it. “Sir?” A girl carrying a basket of thread pointed toward the far wall. “Why is that door bolted?” Otto looked from the lamp to the heavy iron bar. Forty seconds. He could not do both. He ran to the door, heaved the bar free, and dropped it. Behind him, someone shouted. The lamp struck the floor. “Outside!” Otto called. Workers poured through the open door. The girl caught a younger boy by the hand and pulled him along. As Otto stepped aside for them, his minute ended. He returned to find the fire still recorded in the history book. Beneath the date was a sentence that had not been there before: Everyone escaped.
344 tokens in, 696 out (489 of them reasoning) · 9.1 s · $0.0038 · 1 message on Pro · answered by openai/gpt-6-sol via OpenAI ·
Claude Sonnet 5
Passed: Graded 4.3 of 5 on average (lowest 4).
The Minute of 1900 Otto's boots hit cobblestone with a hiss of ozone. One minute. That was all the machine could hold in 1900 before it yanked him back. He'd rehearsed this a hundred times: find the letter, burn it, prevent the war that would come. But the street was wrong—wrong shop signs, wrong faces. Forty seconds left. A boy chased a hoop past him, laughing. An old woman dropped her basket of apples. Otto knelt, gathered them, pressed them into her arms. She blinked at his strange coat and said, "Bless you, stranger." Twenty seconds. He never found the letter. Ten seconds. The boy caught his hoop and grinned at Otto like they shared a secret. The machine seized him back into white light. Otto never learned what changed. But somewhere, an old woman told her grandchildren about a strange kind man, and that story outlived every letter he might have burned.
449 tokens in, 324 out (15 of them reasoning) · 5.9 s · $0.0041 · 1 message on Pro · answered by anthropic/claude-sonnet-5 via Anthropic ·
Grok 4.7
Passed: Graded 4.7 of 5 on average (lowest 4).
Sixty Seconds in 1900 Otto's watch struck, and the street of 1900 snapped into place. He had one minute. One change. A boy tugged a cart toward the river. Papers slid into the mud. Otto had meant to stop a factory fire down the lane. "Help me," the boy said. The factory bell rang. Smoke would start if Otto ran now. Half his minute was already gone. He knelt, gathered the wet pages, and pressed them into the boy's hands. "Take the high road. Don't look back." The boy ran. Otto's watch clicked to zero, and the street vanished. In his own year, a letter waited on his desk, signed by the name he had pulled from the mud. The factory had burned. The boy had lived to write the law that kept the next one from burning at all.
1,578 tokens in, 826 out (651 of them reasoning) · 12.7 s · $0.0046 · 1 message on Pro · answered by x-ai/grok-4.7 via xAI ·
GPT-6 Astra
Passed: Graded 4.7 of 5 on average (lowest 4).
The Minute Hand Otto landed in the year 1900 with sixty seconds glowing on his wrist. He could change one thing. After that, history would harden around him and send him home. Across the street, a clockmaker was locking his shop. Beside the door stood a little girl holding a broken pocket watch. “Tomorrow,” the clockmaker said. Otto knew that watch. It sat beneath glass in his own century, beside a photograph of the girl who had become his grandmother. She had never heard her father’s last message, hidden inside its case. Forty seconds. “Please,” Otto told the clockmaker. “Open it.” “Come back tomorrow.” Twenty seconds. Otto took the watch and pressed its stubborn catch against the doorframe. The lid sprang loose. A folded scrap fluttered out. The girl read it. “He wasn’t leaving us,” she whispered. “He wanted us to follow.” Otto vanished. At home, the museum photograph showed three people. He stayed until closing.
344 tokens in, 213 out · 29.8 s · $0.0070 · 1 message on Pro · answered by openai/gpt-6-astra via OpenAI ·
Gemini 3.1 Pro
Passed: Graded 4.3 of 5 on average (lowest 3).
Sixty Seconds in 1900 Otto materialized on a bustling cobblestone street in 1900, his pocket watch ticking loudly. He had exactly one minute to change one thing. Fifty seconds. He pushed past women in heavy dresses and men in bowler hats, scanning the brick storefronts. Thirty seconds. There it was, the red iron mail box. A young man stood before it, holding a sealed blue envelope. That single letter would ruin Otto's family for three generations. Fifteen seconds. The young man extended his arm toward the brass slot. Otto lunged. Pardon me, he yelled, slamming heavily into the man's shoulder. The blue envelope slipped from the man's fingers, fluttering down into a deep, muddy street puddle. The man groaned, kneeling to retrieve the ruined, soaked paper. Shaking his head, he crumpled it up and walked away. Two seconds. Otto smiled as the antique street dissolved into white light. When he opened his eyes in his own time, his family's dilapidated apartment had become a sprawling, sunlit mansion.
361 tokens in, 1,346 out (1,112 of them reasoning) · 12.0 s · $0.0084 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google AI Studio ·
Claude Opus 5.5
Passed: Graded 4.7 of 5 on average (lowest 4).
Sixty Seconds in 1900 Otto landed in a muddy street as his wrist dial began its countdown: 0:59. He had rehearsed this for years. Find the clockmaker's shop. Stop the fire. Save the town. 0:48. The shop stood untouched, its windows glowing. No smoke. His calculations were wrong. "Mister, you dropped this," said a small boy, holding out Otto's notebook. 0:31. Otto looked past him. A tired lamplighter was stumbling toward the shop, lantern swinging, eyes half closed. "What's your name?" Otto asked. "Felix." 0:19. Otto pressed the notebook back into the boy's hands. "Keep it. Now run and tell that man to put his lantern down. Hurry." Felix ran. The lamplighter blinked, yawned, and set the lantern safely on a stone step. 0:03. The world folded. Otto woke in his laboratory. On the wall hung a photograph he had never seen: the clockmaker's shop, still standing, with a brass plaque reading "Felix & Sons, Since 1900."
451 tokens in, 926 out (553 of them reasoning) · 10.6 s · $0.0203 · 1 message on Pro · answered by anthropic/claude-opus-5.5 via Google ·
Claude Fable 5.1
Passed: Graded 4.7 of 5 on average (lowest 4).
One Minute, Paris, 1900 The machine spat Otto onto a cobbled street, and the wristband began counting down from sixty. He had rehearsed this for years. Turn left, twenty paces, the café where the young engineer would drop his notebook in the gutter and lose the design that might have made the century kinder. Fifty seconds. A boy stood in the road, staring up at a runaway cart horse thundering toward him. Nobody else had noticed yet. Otto froze. The notebook, or the boy. Thirty seconds. He ran the wrong way, seized the child by the collar, and hauled him onto the pavement as the cart crashed past. "Merci, monsieur," the boy gasped. Ten seconds. The café was too far now. "What's your name?" Otto asked. "Henri." The wristband flashed zero. Otto vanished. Back in the lab, the archive had changed. No famous notebook. But a new entry: Henri, engineer, 1900 to 1978.
451 tokens in, 335 out · 7.0 s · $0.0213 · 1 message on Pro · answered by anthropic/claude-fable-5.1 via Google ·
More AI story generator test inputs
The other AI story generator test inputs, each with every model's reply, and the results across all five.
Questions
Which AI does best on “A whole story in 150 words”?
13 of 16 models passed this AI story generator test input's check (rubric); Claude Sonnet 5.5, GPT-6 Luna, and DeepSeek V4 Pro didn't. The cheapest reply that passed was GLM 5.3 Flash's, at $0.00016; the fastest, GLM 5.3's in 3.1 s. The dearest reply, Claude Fable 5.1's, cost 134 times as much ($0.0213).
What does a reply to “A whole story in 150 words” cost?
Through the models' APIs, what OpenRouter charged us ran from $0.00013 (GPT-6 Luna) to $0.0213 (Claude Fable 5.1) for this test input. The cheapest that passed was GLM 5.3 Flash's, $0.00016. In llmwise you don't pay by the token: a reply like these counts as one message on Pro, whichever model answers.
Claude, GPT, Gemini, DeepSeek, Grok, Kimi, and GLM, in one chat.
See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.