claude · anthropic · chat & research
Claude, without a separate Claude plan.
Anthropic's models for long documents, careful writing and code, in the same workspace as every other model you use.
- made by
- Anthropic
- versions
- 5
- from
- 1 token
- included on
- Starter+
best at
What Claude is good at
Claude earns its reputation on work that rewards patience: reading a 90-page contract and finding the clause that contradicts page 12, rewriting a draft without sanding off the author's voice, or following a ten-point style guide to the letter. It tends to say when it isn't sure, which matters more than it sounds when the answer goes into a report.
For code, Claude is strongest when it can see a lot at once. Paste several files, describe the bug and the behaviour you expected, and it will usually trace the problem across them rather than patching the first suspicious line.
On LazySusan you pick the Claude version per conversation, and your chats sit next to the ones you've had with ChatGPT, Gemini and the rest, so you can take a Claude draft to another model for a second opinion without copying anything between apps.
versions
Claude versions on LazySusan
| Version | Use it for |
|---|---|
| Claude Auto | Picks for you: Haiku 5.5 for short questions, Sonnet 5.5 for everything else, including files and long prompts. |
| Claude Fable 5.1 | Anthropic's most capable model. Worth it for the hardest analysis and long, multi-step reasoning. |
| Claude Opus 5.5 | Deep analysis, long documents and serious code at a lower cost than Fable. |
| Claude Sonnet 5.5 | The everyday choice: strong writing and code, quick enough for back-and-forth. |
| Claude Haiku 5.5 | The fastest and cheapest Claude, for quick answers, summaries and rewrites. |
pricing
What Claude costs on LazySusan
| Generation | Tokens |
|---|---|
| Claude Haiku 5.51–2 from a short to a long reply | 1 |
| Claude Sonnet 5.57–40 from a short to a long reply | 16 |
| Claude Opus 5.514–79 from a short to a long reply | 31 |
| Claude Fable 5.135–197 from a short to a long reply | 77 |
Costs are for a message of about 300 characters; you're billed for what the reply actually uses. A typical Sonnet 5.5 reply is 16 tokens, so Starter's 1,000 monthly tokens cover roughly 62 of them, or many more on Haiku.
Every plan includes Claude. Starter is $8 a month with 1,000 tokens; unused tokens roll over while you're subscribed. Compare plans.
prompts
Prompts to try with Claude
// contracts
Here is a supplier agreement. List every obligation that falls on us, with the clause number, and flag anything that conflicts with another clause.
// editing
Tighten this essay by about 20% without changing its argument or its voice. Show me the three cuts you're least sure about.
// code review
These three files handle checkout. A discount is sometimes applied twice. Find where and explain the sequence that causes it.
// research synthesis
Summarise these five interview transcripts into themes. Quote the line that best supports each theme and note where people disagreed.
// style guide
Rewrite these product descriptions to follow the attached style guide. After each one, list the rules you applied.
tips
Getting better results
- 01Give Claude the whole document rather than an excerpt. It does its best work with full context and will point to page or clause numbers if you ask.
- 02Ask for a plan before the answer on long tasks ("outline your approach first"). You can correct the plan cheaply before it writes 2,000 words.
- 03Start on Sonnet 5.5 and step up to Opus 5.5 or Fable 5.1 only when the answer isn't good enough. A typical Opus reply is 31 tokens and Fable 77 tokens, against 16 tokens for Sonnet.
- 04Use Haiku 5.5 for mechanical jobs such as reformatting, translating short text or pulling fields out of an email. At around 1 token a reply it is close to free.
alternatives
When to pick a different tool
- ChatGPT if you want quick, confident answers for everyday tasks, or images in the same conversation.
- Gemini if you're working with very large files, video or images and want speed over polish.
- Perplexity if the question depends on what's on the web today and you need sources you can click.
- Multi-Chat if the answer matters enough to see how Claude, ChatGPT and Gemini each handle it side by side.
also on the api
Claude in your own app
Call every Claude model with one key and the same request shape you use for GPT-6 or Gemini.
One key, our SDKs for JavaScript and Python, prepaid credits with no subscription. LazySusan API · docs
faq
Claude on LazySusan: questions
Yes. Every LazySusan plan includes Claude, from Starter at $8 a month. You don't need an Anthropic account, and the same plan covers ChatGPT, Gemini, image, video and voice models.
Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5, plus Claude Auto, which picks Haiku for short questions and Sonnet for the rest. You can switch version in any conversation.
It's metered in tokens from your plan. A typical message of a few hundred characters costs about 1 token on Haiku 5.5, 16 tokens on Sonnet 5.5, 31 tokens on Opus 5.5 and 77 tokens on Fable 5.1. Long replies and big attachments cost more.
Yes. Attach files to a Claude conversation and ask about them. Claude reads them along with your message, and the cost of reading them counts toward that message.
Yes. You can import past Claude and ChatGPT conversations into LazySusan so they're searchable alongside your new chats.
Yes. All four Claude models are in the developer API, with one key and the LazySusan SDK for JavaScript and Python. The API uses its own prepaid balance, separate from app plans.
get started
Claude and every other model, one plan.
One login, one bill and one Library for everything you make, with the right model for each part of the job.
Starter is $8/month. Cancel anytime.