Fable special: what changes with Claude's new model and how to put it to work
The short version. Anthropic released Claude Fable 5.1 on September 1, 2026 and describes it as its most capable model open to all customers (official guide to choosing a model). It is built for long, demanding work: sessions that run for hours, multistep research, and analysis carried through to a finished document, spreadsheet or slide deck. It is available to Pro, Max, Team and Enterprise users, according to the model page.
The Claude lineup as of October 14, 2026
Anthropic has refreshed the whole family in six weeks. This is how its official models table reads today:
- Fable 5.1 (September 1): demanding reasoning and long-horizon agentic work. The slowest of the four.
- Opus 5.5 (September 22): long-running coding and knowledge work. The one Anthropic recommends starting with for most workloads.
- Sonnet 5.5 (September 28): the balance of speed and intelligence.
- Haiku 5.5 (October 7): the fastest and cheapest, for high-volume tasks.
There is also Mythos 5.1, with the same capabilities as Fable 5.1, offered only to organizations verified through Anthropic's programs.
The detail most write-ups skip
Anthropic's documentation does not say "always use Fable". It says the opposite: start with Opus 5.5 and move to Fable 5.1 when the work calls for hard reasoning or many hours in a row, or when Opus falls short. Its release notes add that Opus 5.5 performs at the level of Fable 5.1 on most work. Monday's email does not need the big model. The report that takes you two days is worth a try.
Where Fable 5.1 improves
According to the official list of what's new, the gains over Fable 5 concentrate in six areas: coding over long sessions, work with documents, spreadsheets and slides, research and web search, reading charts and tables inside PDFs, very long contexts (up to one million tokens, roughly 555,000 words), and operating a browser and a computer. Anthropic says multilingual performance is on par with Fable 5.
Four practical uses, each with its brief
The examples are ours. Swap in your own details.
1. A report built from many documents
"I'm attaching the 14 sets of minutes from our residents' association for 2025 and 2026. I want a two-page report for the board: what was agreed, what is still pending and who is responsible for each item. If two sets of minutes contradict each other, tell me and give the date of each. Do not treat anything as agreed unless a vote is recorded."
2. A spreadsheet you can keep using
"From these three bank statements, build a sheet with spending by category and month. Use formulas, not pasted numbers, so it updates when I add October. Highlight in yellow any charge you cannot classify instead of inventing a category for it."
3. Research with sources
"Compare three invoicing tools for a freelancer in Spain. Use only each vendor's official pages and the tax agency's site. Give me a table with the link for every fact and say clearly what you could not confirm."
4. A dense PDF with tables and charts
"This is the specification for a public tender (86 pages). Tell me the solvency requirements, the deadlines and the scoring criteria, with the page number for each. When you copy a sentence from the document, put it in quotation marks."
That last sentence is there for a reason: in its prompting guide for Fable 5.1, Anthropic notes that when summarizing, this model is more likely to reproduce passages of the source without marking them as quotations.
What to know before using it for everything
- It is slower. The official table lists it as the slowest model in the lineup.
- Some requests go to another model. Queries that its safeguards flag as cybersecurity or biology are automatically routed to an Opus model, according to the Fable page.
- The text carries a watermark. It is statistical and invisible. Anthropic states that it does not change the content and carries no information about you.
- It costs more on the API. $10 per million input tokens and $50 per million output tokens, against $4 and $20 for Opus 5.5 (official models table).
What does not change
The four briefs above look alike: each says who the result is for, what material there is, what shape it should take, what to do with doubts and how to check it. That is briefing well, and it works the same with Fable, with Opus or with whatever model arrives next spring. Anthropic confirms it in its own way: prompts that worked on Fable 5 should work on Fable 5.1 without changes.
A better model can carry more work, but it cannot guess what you have not told it. People who learn to explain the job do not start from zero with every release.
Facts checked on October 14, 2026 against the Anthropic pages linked in the text. Models and their terms change: when in doubt, the official page wins.
The free guide with 10 professional prompts for Claude is here (in Spanish).