Claude Fable 5.2 · Writing
Claude Fable 5.2 for writing: our test runs
Claude Fable 5.2 ran our 5 writing prompts in llmwise, through the same pipeline your messages take. Here's every reply as it came, whether it passed, what it cost and what it counts as on Pro, next to the other models.
Based on 5 of our test runs on , through OpenRouter with the app's own prompt and settings.
Short answer
In our test runs on November 17, 2026, Claude Fable 5.2 passed 5 of 5 of our writing prompts (both hard ones): tied 1st of 27 models. Each reply counted as 1 message on Pro, where it has 31 a month, and cost $0.0115 to run on average.
Prompt by prompt
Each of the 5 writing prompts, and how Claude Fable 5.2's reply did.
| Prompt | Result | Cost | Time |
|---|---|---|---|
| Announce a second bakery shop on LinkedInEveryday | Passed: Graded 4.5 of 5 on average (lowest 4). | $0.0176 | 4.6 s |
| Rewrite corporate jargon in plain wordsEveryday | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0059 | 2.5 s |
| Decline a meeting and offer two timesEveryday | Passed: Graded 5.0 of 5 on average (lowest 5). | $0.0081 | 3.3 s |
| A product announcement with five rulesHard | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0102 | 3.5 s |
| Argue both sides of free busesHard | Passed: Graded 4.7 of 5 on average (lowest 4). | $0.0155 | 5.4 s |
Every writing reply from Claude Fable 5.2
Announce a second bakery shop on LinkedIn
An everyday prompt: “Announce a second bakery shop on LinkedIn”, in full.
Passed: Graded 4.5 of 5 on average (lowest 4).
We have some news we've been itching to share: Crumb & Co. is opening a second shop, and it's in Leeds. Doors open on 14 November. When we started, it was one small oven, a handful of recipes and a lot of early mornings. We never imagined we'd outgrow our first counter, but thanks to everyone who has queued, chatted and come back week after week, here we are. The new shop will have everything you already know us for: fresh bread baked each morning, proper pastries and cakes made by hand. Same bakers, same recipes, just a little more room. To say thank you, coffee is on us all day on opening day. Come in, say hello and have a look around. We're so grateful to everyone who helped us get here. See you in Leeds on the 14th.
794 tokens in, 244 out · 4.6 s · $0.0176 · 1 message on Pro · answered by anthropic/claude-fable-5.2 via Google ·
Rewrite corporate jargon in plain words
An everyday prompt: “Rewrite corporate jargon in plain words”, in full.
Passed: Graded 4.7 of 5 on average (lowest 4).
In the third quarter, our teams worked together on a project that put customers first. As a result, customer retention measures improved by 12%, and the rate at which we lose large-business customers slowed by 0.3 percentage points. We also scaled back the number of channels we use to reach customers so our staff could make better use of their time.
853 tokens in, 99 out · 2.5 s · $0.0059 · 1 message on Pro · answered by anthropic/claude-fable-5.2 via Google ·
Decline a meeting and offer two times
An everyday prompt: “Decline a meeting and offer two times”, in full.
Passed: Graded 5.0 of 5 on average (lowest 5).
**Subject: Vendor demo – alternative times** Hi Tom, Thanks for the invitation to the vendor demo on Wednesday at 11:00. Unfortunately, I'm presenting to the board at that time and won't be able to attend. Would either of these work instead? - Thursday at 10:00 - Friday at 14:00 If neither suits, let me know and I'll find another slot. Sorry for the inconvenience, and thanks for understanding. Best regards, [Your name]
798 tokens in, 167 out · 3.3 s · $0.0081 · 1 message on Pro · answered by anthropic/claude-fable-5.2 via Google ·
A product announcement with five rules
A hard prompt: “A product announcement with five rules”, in full.
Passed: Graded 4.7 of 5 on average (lowest 4).
**Meet the Nordvik 2** Good light should adapt to you, not the other way around. The Nordvik 2 desk lamp dims in five steps, taking you from bright, focused task lighting to a soft evening glow with a single touch. Its arm turns 270 degrees, so the light lands exactly where you need it, whether that's a keyboard, a sketchbook, or the page of a novel. It charges from USB-C, the same cable you already use for everything else, keeping your desk free of extra clutter. The Nordvik 2 costs €79 and ships on 1 December, just in time for the darkest evenings of the year. Where will yours sit?
810 tokens in, 215 out · 3.5 s · $0.0102 · 1 message on Pro · answered by anthropic/claude-fable-5.2 via Google ·
Argue both sides of free buses
A hard prompt: “Argue both sides of free buses”, in full.
Passed: Graded 4.7 of 5 on average (lowest 4).
Free buses turn transit into what it should be: basic infrastructure, like sidewalks and streetlights. Eliminating fares removes a regressive cost that falls hardest on low-income riders, speeds boarding, and ends costly fare collection and enforcement. Ridership rises, pulling cars off congested roads and cutting emissions. Fares often cover only a small share of operating costs anyway, so the lost revenue is modest compared with the gains: wider access to jobs, schools, and healthcare, and a city where mobility doesn't depend on what's in your wallet. Riders consistently say they want frequent, reliable, fast service more than cheap service, and free fares drain the very revenue that could fund it. Evidence from cities that tried it shows most new trips come from people who previously walked or cycled, not drivers, so congestion and emissions barely move. A system wholly dependent on political budgets becomes vulnerable to cuts in every downturn. Targeted discounts for low-income riders deliver equity at a fraction of the cost, leaving money to run more buses, which is what actually grows ridership.
774 tokens in, 358 out · 5.4 s · $0.0155 · 1 message on Pro · answered by anthropic/claude-fable-5.2 via Google ·
How it compares on writing
Claude Fable 5.2 beside the job's picks among every model. 14 models share the hard-problems pick, Claude Fable 5.2 among them.
| Model | Passed | On Pro | Cost per reply |
|---|---|---|---|
| Gemini 3.8 FlashOur pick: hard problems (shared), best value | 5 of 5 | 250 a month on Pro | $0.0005 |
| DeepSeek V4.1 FlashOur pick: everyday | 4 of 5 | 60 a day on Pro | $0.0003 |
| Claude Fable 5.2Our pick: hard problems (shared) | 5 of 5 | 31 a month on Pro | $0.0115 |
Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Plus $3.75, Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.
Claude Fable 5.2 next to each model on writing
Claude Fable 5.2 against each model on the same 5 writing prompts. One run each, through OpenRouter: a wait depends on the provider and the load that day, so a lead under 10% counts as close.
Against Claude Fable 5.1: Claude Fable 5.2 answered 1.2× sooner (3.5 s to 4.1 s), cost 1.2× less, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Argue both sides of free buses.
Against Claude Opus 5.5: Claude Fable 5.2 answered 1.4× sooner (3.5 s to 4.8 s), cost 1.4× less, and passed 5 of 5 to its 5.
Against Claude Sonnet 5.5: Claude Fable 5.2 answered 1.3× later (3.5 s to 2.7 s), cost 4.4× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Argue both sides of free buses.
Against Claude Sonnet 5: Claude Fable 5.2 answered within 10% of its time (3.5 s to 3.6 s), cost 3.0× more, and passed 5 of 5 to its 3. Only Claude Fable 5.2 passed Announce a second bakery shop on LinkedIn; A product announcement with five rules.
Against Claude Haiku 5.5: Claude Fable 5.2 answered 2.1× later (3.5 s to 1.7 s), cost 86.4× more, and passed 5 of 5 to its 3. Only Claude Fable 5.2 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Claude Haiku 4.5: Claude Fable 5.2 answered 2.1× later (3.5 s to 1.7 s), cost 8.8× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed A product announcement with five rules.
Against GPT-6.2 Astra: Claude Fable 5.2 answered 1.4× sooner (3.5 s to 4.9 s), cost 1.3× more, and passed 5 of 5 to its 5.
Against GPT-6 Astra: Claude Fable 5.2 answered 1.3× sooner (3.5 s to 4.7 s), cost 1.9× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Rewrite corporate jargon in plain words.
Against GPT-6.2 Sol: Claude Fable 5.2 answered 1.9× later (3.5 s to 1.8 s), cost 5.2× more, and passed 5 of 5 to its 5.
Against GPT-6.1 Sol: Claude Fable 5.2 answered 1.6× sooner (3.5 s to 5.5 s), cost 6.0× more, and passed 5 of 5 to its 5.
Against GPT-6 Sol: Claude Fable 5.2 answered within 10% of its time (3.5 s to 3.3 s), cost 8.6× more, and passed 5 of 5 to its 5.
Against GPT-6.1 Terra: Claude Fable 5.2 answered 1.1× sooner (3.5 s to 3.9 s), cost 17.5× more, and passed 5 of 5 to its 5.
Against GPT-6 Luna: Claude Fable 5.2 answered 1.3× later (3.5 s to 2.6 s), cost 204.7× more, and passed 5 of 5 to its 3. Only Claude Fable 5.2 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Gemini 4 Carbon: Claude Fable 5.2 answered 6.0× sooner (3.5 s to 21.0 s), cost 1.8× more, and passed 5 of 5 to its 5.
Against Gemini 4 Argon: Claude Fable 5.2 answered 2.3× sooner (3.5 s to 7.9 s), cost 8.2× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Argue both sides of free buses.
Against Gemini 4 Hydrogen: Claude Fable 5.2 answered 2.3× sooner (3.5 s to 7.9 s), cost 1.7× more, and passed 5 of 5 to its 5.
Against Gemini 3.1 Pro (preview): Claude Fable 5.2 answered 2.1× sooner (3.5 s to 7.4 s), cost 2.3× more, and passed 5 of 5 to its 3. Only Claude Fable 5.2 passed A product announcement with five rules; Argue both sides of free buses.
Against Gemini 3.8 Flash: Claude Fable 5.2 answered 1.1× later (3.5 s to 3.1 s), cost 23.3× more, and passed 5 of 5 to its 5.
Against DeepSeek V4.1 Flash: Claude Fable 5.2 answered 2.2× later (3.5 s to 1.6 s), cost 34.9× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Rewrite corporate jargon in plain words.
Against DeepSeek V4 Pro: Claude Fable 5.2 answered within 10% of its time (3.5 s to 3.3 s), cost 8.9× more, and passed 5 of 5 to its 5.
Against Grok 4.7: Claude Fable 5.2 answered 2.4× sooner (3.5 s to 8.6 s), cost 2.0× more, and passed 5 of 5 to its 5.
Against Kimi K3.1 Pro: Claude Fable 5.2 answered 2.0× sooner (3.5 s to 6.9 s), cost 2.3× more, and passed 5 of 5 to its 5.
Against Kimi K3: Claude Fable 5.2 answered 1.5× sooner (3.5 s to 5.1 s), cost 3.5× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Argue both sides of free buses.
Against GLM 5.3: Claude Fable 5.2 answered 2.3× later (3.5 s to 1.5 s), cost 13.3× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed Argue both sides of free buses.
Against GLM 5.3 Flash: Claude Fable 5.2 answered 3.4× later (3.5 s to 1.0 s), cost 70.2× more, and passed 5 of 5 to its 4. Only Claude Fable 5.2 passed A product announcement with five rules.
Against Mistral Large 4: Claude Fable 5.2 answered 2.1× sooner (3.5 s to 7.2 s), cost 7.8× more, and passed 5 of 5 to its 5.
How these runs were done
Rubric (graded). The grader model scores the reply from 1 to 5 on each published criterion. It passes with an average of 4 or more and no criterion under 3, and only if it also meets the prompt's automatic rules (length, words it must or mustn't use).
The grader is Claude Opus 5.5 at low reasoning effort; its own replies are graded by GPT-6 Astra, so no model grades itself. Its prompt and every rubric are on the methods page.
Claude Fable 5.2, and writing, elsewhere
- Claude Fable 5.2: price, limits and messages on every plan
- The best AI for writing: our picks among every model
- Claude for writing
- Claude Fable 5.2 for coding: our test runs
- Claude Fable 5.2 for math: our test runs
- Claude Fable 5.2 for data analysis: our test runs
- Rewrite for clarity: the prompt, and its price on each model
- AI prompts for writing, tested on four models
- Essay prompts for AI that help you write your own
- Our test runs: 50 prompts on every model
Questions
Is Claude Fable 5.2 good for writing?
In our test runs it passed 5 of 5 writing prompts, tied 1st of the 27 models in llmwise. Every reply is on this page, so you can judge them yourself.
How many of my messages does a writing reply from Claude Fable 5.2 use?
1 message each on Pro, where it has 31 a month on Pro. The price of a message is fixed and shown before you send it, however long the reply.
How were these runs done?
The same way for every model: each prompt sent through llmwise's own pipeline, each reply checked the same way. The methods page has every prompt and how each is scored.
Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral, in one chat.
See what a message costs before you send it. Every plan is month to month: cancel in two clicks. Sign in with an email link, no password.