Claude Haiku 5.5 · Writing
Claude Haiku 5.5 for writing: our test runs
Claude Haiku 5.5 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 18, 2026, Claude Haiku 5.5 passed 3 of 5 of our writing prompts (1 of the 2 hard ones): tied 24th of 27 models. Each reply counted as 1 message on Pro, where it has 60 a day, and cost $0.0001 to run on average.
Prompt by prompt
Each of the 5 writing prompts, and how Claude Haiku 5.5's reply did.
| Prompt | Result | Cost | Time |
|---|---|---|---|
| Announce a second bakery shop on LinkedInEveryday | Passed: Graded 4.8 of 5 on average (lowest 4). | $0.0001 | 1.8 s |
| Rewrite corporate jargon in plain wordsEveryday | Failed: Graded 3.7 of 5 on average (lowest 3). | $0.0001 | 1.1 s |
| Decline a meeting and offer two timesEveryday | Passed: Graded 5.0 of 5 on average (lowest 5). | $0.0001 | 1.3 s |
| A product announcement with five rulesHard | Passed: Graded 4.0 of 5 on average (lowest 3). | $0.0001 | 1.7 s |
| Argue both sides of free busesHard | Failed: Graded 3.3 of 5 on average (lowest 2); but a paragraph of 116 words, over the 90 allowed. | $0.0003 | 3.6 s |
Every writing reply from Claude Haiku 5.5
Announce a second bakery shop on LinkedIn
An everyday prompt: “Announce a second bakery shop on LinkedIn”, in full.
Passed: Graded 4.8 of 5 on average (lowest 4).
Big news from Crumb & Co.: we're opening a second shop in Leeds on 14 November! When we started, we never imagined our little bakery would grow beyond one counter and one oven. But you kept coming back, bringing friends, and telling us what you loved. Now we're bringing the same warm bread, slow-proved sourdough and buttery pastries to a new home in Leeds, and we can't wait to meet you there. To celebrate, every guest on opening day gets a free coffee. Pop in, say hello, and let us treat you. Thank you for being part of our story. We'll see you on the 14th.
795 tokens in, 199 out · 1.8 s · $0.0001 · 1 message on Pro · answered by anthropic/claude-haiku-5.5 via Anthropic ·
Rewrite corporate jargon in plain words
An everyday prompt: “Rewrite corporate jargon in plain words”, in full.
Failed: Graded 3.7 of 5 on average (lowest 3).
The Q3 initiative, built on teamwork across departments, changed how we serve customers. Retention rose 12%, and customers left at a rate 30 basis points lower across our enterprise segment. We also streamlined our multichannel contact points to use resources more efficiently.
854 tokens in, 88 out · 1.1 s · $0.0001 · 1 message on Pro · answered by anthropic/claude-haiku-5.5 via Anthropic ·
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: Re: Vendor Demo on Wednesday Hi Tom, Thank you for the invitation to the vendor demo on Wednesday at 11:00. Unfortunately, I'm presenting to the board at that time, so I won't be able to attend. Would either of these work instead? - Thursday at 10:00 - Friday at 14:00 Please let me know which suits you best, and I'll make it work. Best regards, [Your Name]
799 tokens in, 153 out · 1.3 s · $0.0001 · 1 message on Pro · answered by anthropic/claude-haiku-5.5 via Anthropic ·
A product announcement with five rules
A hard prompt: “A product announcement with five rules”, in full.
Passed: Graded 4.0 of 5 on average (lowest 3).
Introducing the Nordvik 2 desk lamp. Its five-step dimmer lets you pick the exact brightness for reading, writing, or winding down. Charge it with a standard USB-C cable, so one port keeps it running without hunting for a special adapter. The arm turns a full 270 degrees, letting you aim light precisely where you need it. The Nordvik 2 costs €79 and ships on 1 December. Pre-order yours now and light up your desk the way you want. Will you be the first on your street to upgrade your workspace?
811 tokens in, 176 out · 1.7 s · $0.0001 · 1 message on Pro · answered by anthropic/claude-haiku-5.5 via Anthropic ·
Argue both sides of free buses
A hard prompt: “Argue both sides of free buses”, in full.
Failed: Graded 3.3 of 5 on average (lowest 2); but a paragraph of 116 words, over the 90 allowed.
Making city buses free would boost fairness and economic opportunity. Fares often fall hardest on low-income residents, who may spend a significant share of their income on rides to work, school, and medical appointments. Removing fares eliminates that barrier, expands access to jobs, and keeps people from skipping necessary trips to save money. Free service also speeds boarding, since passengers no longer queue to pay, which can improve reliability and reduce delays. Ridership typically rises, cutting traffic congestion and emissions. Fare collection itself is costly, from equipment and maintenance to enforcement, so eliminating it can free money for service expansion. Cities that have tried free transit report higher usage and stronger public support for the system. Free buses would divert scarce public money from needs that matter more. Fare revenue, though often a modest share of operating costs, funds routes, drivers, and maintenance, and losing it forces cities to raise taxes or cut service elsewhere. Riders who pay fares also have a stake in the system and tend to demand quality, while free service can invite overcrowding, vandalism, and safety problems that drive away paying riders. Fares also encourage efficient use: when transit costs nothing, demand can outstrip capacity on busy routes. Wealthier residents, who already own cars or use rail, may benefit disproportionately. Targeted discounts for low-income riders deliver relief more precisely than a blanket policy that subsidizes everyone equally.
775 tokens in, 472 out · 3.6 s · $0.0003 · 1 message on Pro · answered by anthropic/claude-haiku-5.5 via Anthropic ·
How it compares on writing
Claude Haiku 5.5 beside the job's picks among every model. 14 models share the hard-problems pick, and Claude Haiku 5.5 isn't one of 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 Haiku 5.5 | 3 of 5 | 60 a day on Pro | $0.0001 |
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 Haiku 5.5 next to each model on writing
Claude Haiku 5.5 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 5.2: Claude Haiku 5.5 answered 2.1× sooner (1.7 s to 3.5 s), cost 86.4× less, and passed 3 of 5 to its 5. Only Claude Fable 5.2 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Claude Fable 5.1: Claude Haiku 5.5 answered 2.4× sooner (1.7 s to 4.1 s), cost 100.9× less, and passed 3 of 5 to its 4. Only Claude Fable 5.1 passed Rewrite corporate jargon in plain words.
Against Claude Opus 5.5: Claude Haiku 5.5 answered 2.9× sooner (1.7 s to 4.8 s), cost 119.8× less, and passed 3 of 5 to its 5. Only Claude Opus 5.5 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Claude Sonnet 5.5: Claude Haiku 5.5 answered 1.6× sooner (1.7 s to 2.7 s), cost 19.7× less, and passed 3 of 5 to its 4. Only Claude Sonnet 5.5 passed Rewrite corporate jargon in plain words.
Against Claude Sonnet 5: Claude Haiku 5.5 answered 2.1× sooner (1.7 s to 3.6 s), cost 28.6× less, and passed 3 of 5 to its 3. Only Claude Haiku 5.5 passed Announce a second bakery shop on LinkedIn; A product announcement with five rules. Only Claude Sonnet 5 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Claude Haiku 4.5: Claude Haiku 5.5 answered within 10% of its time (1.7 s to 1.7 s), cost 9.8× less, and passed 3 of 5 to its 4. Only Claude Haiku 5.5 passed A product announcement with five rules. Only Claude Haiku 4.5 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6.2 Astra: Claude Haiku 5.5 answered 2.9× sooner (1.7 s to 4.9 s), cost 65.8× less, and passed 3 of 5 to its 5. Only GPT-6.2 Astra passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6 Astra: Claude Haiku 5.5 answered 2.8× sooner (1.7 s to 4.7 s), cost 46.5× less, and passed 3 of 5 to its 4. Only GPT-6 Astra passed Argue both sides of free buses.
Against GPT-6.2 Sol: Claude Haiku 5.5 answered within 10% of its time (1.7 s to 1.8 s), cost 16.7× less, and passed 3 of 5 to its 5. Only GPT-6.2 Sol passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6.1 Sol: Claude Haiku 5.5 answered 3.2× sooner (1.7 s to 5.5 s), cost 14.4× less, and passed 3 of 5 to its 5. Only GPT-6.1 Sol passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6 Sol: Claude Haiku 5.5 answered 1.9× sooner (1.7 s to 3.3 s), cost 10.0× less, and passed 3 of 5 to its 5. Only GPT-6 Sol passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6.1 Terra: Claude Haiku 5.5 answered 2.3× sooner (1.7 s to 3.9 s), cost 4.9× less, and passed 3 of 5 to its 5. Only GPT-6.1 Terra passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against GPT-6 Luna: Claude Haiku 5.5 answered 1.5× sooner (1.7 s to 2.6 s), cost 2.4× more, and passed 3 of 5 to its 3.
Against Gemini 4 Carbon: Claude Haiku 5.5 answered 12.3× sooner (1.7 s to 21.0 s), cost 47.5× less, and passed 3 of 5 to its 5. Only Gemini 4 Carbon passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Gemini 4 Argon: Claude Haiku 5.5 answered 4.7× sooner (1.7 s to 7.9 s), cost 10.5× less, and passed 3 of 5 to its 4. Only Gemini 4 Argon passed Rewrite corporate jargon in plain words.
Against Gemini 4 Hydrogen: Claude Haiku 5.5 answered 4.6× sooner (1.7 s to 7.9 s), cost 50.6× less, and passed 3 of 5 to its 5. Only Gemini 4 Hydrogen passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Gemini 3.1 Pro (preview): Claude Haiku 5.5 answered 4.4× sooner (1.7 s to 7.4 s), cost 37.7× less, and passed 3 of 5 to its 3. Only Claude Haiku 5.5 passed A product announcement with five rules. Only Gemini 3.1 Pro (preview) passed Rewrite corporate jargon in plain words.
Against Gemini 3.8 Flash: Claude Haiku 5.5 answered 1.8× sooner (1.7 s to 3.1 s), cost 3.7× less, and passed 3 of 5 to its 5. Only Gemini 3.8 Flash passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against DeepSeek V4.1 Flash: Claude Haiku 5.5 answered within 10% of its time (1.7 s to 1.6 s), cost 2.5× less, and passed 3 of 5 to its 4. Only DeepSeek V4.1 Flash passed Argue both sides of free buses.
Against DeepSeek V4 Pro: Claude Haiku 5.5 answered 2.0× sooner (1.7 s to 3.3 s), cost 9.7× less, and passed 3 of 5 to its 5. Only DeepSeek V4 Pro passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Grok 4.7: Claude Haiku 5.5 answered 5.0× sooner (1.7 s to 8.6 s), cost 42.6× less, and passed 3 of 5 to its 5. Only Grok 4.7 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Kimi K3.1 Pro: Claude Haiku 5.5 answered 4.1× sooner (1.7 s to 6.9 s), cost 37.6× less, and passed 3 of 5 to its 5. Only Kimi K3.1 Pro passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Kimi K3: Claude Haiku 5.5 answered 3.0× sooner (1.7 s to 5.1 s), cost 24.4× less, and passed 3 of 5 to its 4. Only Kimi K3 passed Rewrite corporate jargon in plain words.
Against GLM 5.3: Claude Haiku 5.5 answered 1.1× later (1.7 s to 1.5 s), cost 6.5× less, and passed 3 of 5 to its 4. Only GLM 5.3 passed Rewrite corporate jargon in plain words.
Against GLM 5.3 Flash: Claude Haiku 5.5 answered 1.7× later (1.7 s to 1.0 s), cost 1.2× less, and passed 3 of 5 to its 4. Only Claude Haiku 5.5 passed A product announcement with five rules. Only GLM 5.3 Flash passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
Against Mistral Large 4: Claude Haiku 5.5 answered 4.3× sooner (1.7 s to 7.2 s), cost 11.1× less, and passed 3 of 5 to its 5. Only Mistral Large 4 passed Rewrite corporate jargon in plain words; Argue both sides of free buses.
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 Haiku 5.5, and writing, elsewhere
- Claude Haiku 5.5: price, limits and messages on every plan
- The best AI for writing: our picks among every model
- Claude for writing
- Claude Haiku 5.5 for coding: our test runs
- Claude Haiku 5.5 for math: our test runs
- Claude Haiku 5.5 for data analysis: our test runs
- Claude Haiku 5.5 vs GPT-6 Luna
- Claude Haiku 5.5 vs Claude Haiku 4.5
- Claude Sonnet 5.5 vs Claude Haiku 5.5
- Claude Haiku 5.5 vs GLM 5.3 Flash
- Claude Haiku 5.5 vs DeepSeek V4.1 Flash
- 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 Haiku 5.5 good for writing?
In our test runs it passed 3 of 5 writing prompts, tied 24th 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 Haiku 5.5 use?
1 message each on Pro, where it has 60 a day 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.