Sunday, 22 November 2026 · Bengaluru · New Delhi · San Francisco

Claude Fable 5.2: price, plans and when Indian teams should use it

Claude Fable 5.2 costs $8/$40 per million tokens with a 1M context. Plans that include it, cloud availability and rupee costs vs Opus, Sonnet and Haiku.

Share
Claude Fable 5.2: price, plans and when Indian teams should use it
Illustration: 100xFounder

Claude Fable 5.2 is Anthropic’s most capable model open to all customers, released on 9 November 2026 with the API ID claude-fable-5-2, a 1-million-token context window and a price of $8 per million input tokens and $40 per million output tokens, according to Anthropic’s model documentation. That is 20% cheaper than Fable 5.1, but still twice the price of Claude Opus 5.5, which Anthropic recommends as the starting point for most work.

This guide covers what Fable 5.2 is, where Indian teams can use it (Claude apps, the Claude API, AWS, Google Cloud and Microsoft Foundry), the launch offers that ran until 16 November, and how to choose between Fable 5.2 and the Opus, Sonnet and Haiku 5.5 models by cost and task, with rupee examples.

Key takeaways

  • Fable 5.2 launched on 9 November 2026 alongside Claude Mythos 5.2, which Anthropic calls “the same underlying model” with access limited to verified organisations.
  • API prices per million tokens: $8 input, $40 output, $0.20 for cache reads and $4 / $20 through the Batch API, per Anthropic’s pricing docs on 15 November 2026.
  • In the Claude apps, Max ($100 a month) and Ultra ($200) include Fable for up to half the weekly limit; Pro uses paid usage credits for Fable after a launch week that ended on 16 November.
  • Anthropic’s own advice: start on Opus 5.5 and move up to Fable 5.2 only when Opus at high effort still falls short.
  • At ₹94.20 to the dollar, a task with 200,000 input and 30,000 output tokens costs about ₹264 on Fable 5.2, ₹132 on Opus 5.5 and ₹66 on Sonnet 5.5.

How we checked this

Model names, dates, prices and availability below come from Anthropic’s own pages, read on 15 November 2026: the Fable 5.2 and Mythos 5.2 announcement, the Fable 5.2 overview and “What’s new” pages, the models overview, the pricing page and the “Choosing a model” guide on platform.claude.com, the plans page at claude.com/pricing as shown to a visitor in India, and the Claude Help Center article on the Fable 5.2 launch offers. Benchmark figures are Anthropic’s own evaluations. Rupee figures use the RBI reference rate for 13 November 2026, ₹94.1997 to the US dollar, rounded to ₹94.20.

What Claude Fable 5.2 is

Claude Fable 5.2 is built for “demanding reasoning and long-horizon agentic work”, such as agent sessions that run for hours, multistep deep research and analysis carried through to a finished document, spreadsheet or deck, according to its overview page. Anthropic describes it as its most capable model open to all customers.

Anthropic released Fable 5.2 and Claude Mythos 5.2 together on 9 November 2026. Its announcement says the two are the same underlying model: Fable carries stronger cybersecurity and biology safeguards so it can be offered to everyone, while Mythos 5.2 goes only to organisations verified through programmes such as the Cyber Verification Program.

The announcement says Fable 5.2 produces output up to 30% faster than Fable 5.1, In Anthropic’s early testing, Fable 5.2’s safety classifiers blocked about 55% fewer requests than Fable 5.1’s. When a high-risk request is blocked, the user gets an answer from Opus 4.8 for cybersecurity or Opus 5 for biology instead.

Specifications at a glance

  • API ID: claude-fable-5-2 on the Claude API, Claude Platform on AWS, both Google Cloud options and Microsoft Foundry; global.anthropic.claude-fable-5-2 on Amazon Bedrock.
  • Context and output: 1 million tokens of context, 128,000 output tokens; text and images in, text out.
  • Thinking: adaptive thinking is always on; the effort setting (low, medium, high, xhigh, max) controls depth, and the API default is high.
  • Knowledge cutoff: June 2026. Anthropic says it won’t retire the model on its own platforms before 9 November 2027.

Fable 5.2 API price, and how it compares with Opus, Sonnet and Haiku 5.5

Claude Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, according to Anthropic’s pricing documentation, down from $10 and $50 for Fable 5.1. Claude Opus 5.5 costs half as much, Sonnet 5.5 a quarter, and Haiku 5.5 is the cheapest by far.

Claude API prices per million tokens in US dollars, from platform.claude.com pricing and models overview pages, 15 November 2026.

Model Input Output Cache reads Batch (input / output) Speed
Claude Fable 5.2 $8 $40 $0.20 $4 / $20 Moderate
Claude Opus 5.5 $4 $20 $0.20 $2 / $10 Moderate
Claude Sonnet 5.5 $2 $10 $0.10 $1 / $5 Fast
Claude Haiku 5.5, prompts up to 100,000 tokens $0.10 $0.50 $0.01 $0.05 / $0.25 Fastest
Claude Haiku 5.5, prompts over 100,000 tokens $0.50 $2.50 $0.05 $0.25 / $1.25 Fastest

A few pricing details matter more than the headline rates:

  • Cache reads: Fable 5.2 cache reads cost 2.5% of its input price, so at $0.20 per million tokens they cost the same as on Opus 5.5. Prompt-cache writes cost $10 per million tokens for five minutes and $16 for one hour.
  • Fast mode: up to 2.5x faster output at $16 input and $80 output per million tokens. It can’t be combined with the Batch API.
  • Long context: Fable, Opus and Sonnet charge the same per-token rate across the full 1-million-token window. Only Haiku 5.5 charges more once a prompt passes 100,000 tokens.
  • US-only processing: pinning inference to the United States with the inference_geo setting adds a 1.1x multiplier, and regional endpoints on Bedrock and Google Cloud add 10% over global ones.
  • Tokenizer: Claude models from Opus 4.7 onward produce about 30% more tokens for the same text than older Claude models, the pricing page says, so compare costs per task rather than per token when moving off an older model.
Infographic table comparing Claude Fable 5.2, Opus 5.5, Sonnet 5.5 and Haiku 5.5 API prices per million tokens, cache reads, batch rates and best uses
Claude API prices per million tokens, 15 November 2026. · Infographic by 100xFounder

Where Fable 5.2 is available for Indian teams

Claude Fable 5.2 is available to all customers on the Claude API and on AWS, Google Cloud and Microsoft Foundry, according to Anthropic’s “What’s new in Claude Fable 5.2” page. That matters for Indian companies whose cloud spend is already committed to one of those providers.

  • Claude API: all customers, as claude-fable-5-2.
  • Amazon Web Services: through Claude Platform on AWS, or on Amazon Bedrock with the global or US model ID on the bedrock-runtime endpoint only.
  • Google Cloud: both the Anthropic-managed and the Google-managed Claude offerings.
  • Microsoft Foundry: on Anthropic infrastructure.

Two conditions apply. Fable 5.2 requires 30-day data retention and is not available under zero data retention unless Anthropic expressly authorises it, the overview page says. And because Bedrock lists only global and US IDs for the model, teams that need processing in a specific region should confirm the region before they build.

Changes developers moving from Fable 5.1 must make

Anthropic lists four changes from Fable 5.1 that can break existing code. Requests that edit earlier conversation turns fail with a 400 error, other models can’t read Fable 5.2’s thinking blocks, the older computer_20251124 tool errors everywhere except Bedrock, and a Fable 5.2 executor needs a Fable 5.2 or Mythos 5.2 advisor. Forced tool use, through tool_choice “any” or “tool”, also returns a 400 error.

Fable 5.2 in the Claude apps: which plans include it

Claude’s Max plan at $100 a month and Ultra plan at $200 a month include Fable for up to half of the weekly usage limit, according to claude.com/pricing as shown to a visitor in India on 15 November 2026. On Pro, Fable is “paid with usage credits”, and the Free plan does not include it.

Claude plans and Fable access, as shown on claude.com/pricing to a visitor in India on 15 November 2026, in US dollars before applicable tax.

Plan Price Fable access after the launch week
Free $0 Not included
Pro $17 a month billed yearly, $20 monthly Paid with usage credits
Max $100 a month Up to 50% of weekly limits
Ultra $200 a month Up to 50% of weekly limits
Team, standard seat $20 a seat yearly, $25 monthly Usage credits
Team, premium seat $100 a seat yearly, $125 monthly Up to 50% of weekly limits
Enterprise $20 a seat yearly plus usage at API rates, 20 seats minimum Included, billed at API rates

The launch offers ran until 16 November

Pro subscribers and standard Team seats could use up to 50% of their weekly limits on Fable at no extra cost until 16 November 2026 at 11:59 pm Pacific time (1:29 pm IST on 17 November), according to the Claude Help Center’s Fable 5.2 launch offers article. Since then, Fable on those plans runs on usage credits.

  • Pro and standard Team seats also got one free usage reset, which expired at the same time.
  • Max, Ultra and premium Team seats get one to three resets depending on how long they have paid: one under six months, two for six to 12 months and three beyond 12 months. These expire on 8 March 2027.
  • Only accounts with paid billing in place before 10:00 am Pacific time on 9 November 2026 qualify. A reset restores Fable usage on the weekly limit; it doesn’t raise the limit.

How Fable 5.2 scores against Opus 5.5

Claude Fable 5.2 scores 49.1% on the FrontierCode 2.0 agentic coding benchmark at xhigh effort, against 45.5% for Opus 5.5, according to Anthropic’s announcement. Its other reported leads over Opus 5.5 include FrontierSWE v2 (69.9% vs 62.3%), the AutomationBench business-workflow test (51.4% vs 42.5%) and Humanity’s Last Exam with tools (74.2% vs 67.7%).

The same table shows where Fable is not ahead. On OSWorld 2.1 computer use, Fable 5.2 scores 89.3% to Opus 5.5’s 87.2%, a small gap for twice the price. On AutomationBench, Anthropic’s table lists Google’s Gemini 4 Carbon at 53.9%, above Fable 5.2. These are vendor-run tests; your own evaluation on your own tasks is the figure that counts.

Fable, Opus, Sonnet or Haiku: how Indian teams should choose

Anthropic recommends starting with Claude Opus 5.5 for most coding and knowledge work, and moving to Fable 5.2 only when evaluations on Opus 5.5 at xhigh or max effort still fall short, according to its “Choosing the right model” guide. For simpler or high-volume work, it points down to Sonnet 5.5 or Haiku 5.5.

  • Fable 5.2: hours-long agent runs, multistep research, finished reports or decks where a wrong answer costs more than the tokens. It can also act as an advisor that cheaper models consult on hard steps.
  • Opus 5.5: long-running coding agents, large refactors, systems engineering, vision-heavy work and computer use. The default for most product teams.
  • Sonnet 5.5: everyday code generation, data analysis, content and tool-using agents at half Opus’s price.
  • Haiku 5.5: classification, extraction, routing, real-time chat and subagent tasks, where volume decides the bill.

Anthropic’s guide also says tuning effort is often a better lever than switching models. On Fable 5.2, it advises starting at the default high setting and keeping max for rare tasks where time and tokens don’t matter.

What Fable 5.2 costs in rupees: three worked examples

The examples use Anthropic’s list prices on 15 November 2026 and ₹94.20 to the dollar; token counts are our assumptions, and real bills rise with higher effort settings.

1. One long analysis task

A task that reads 200,000 tokens (a large codebase slice or a stack of reports) and writes 30,000 tokens costs about $2.80 (₹264) on Fable 5.2, $1.40 (₹132) on Opus 5.5 and $0.70 (₹66) on Sonnet 5.5. On Haiku 5.5, where the prompt is over 100,000 tokens and the higher rate applies, it costs about $0.18 (₹16).

2. Sorting 1 lakh support tickets a month

At 1,500 input and 100 output tokens per ticket, 1 lakh (100,000) tickets come to 150 million input and 10 million output tokens. That costs about $20 (₹1,884) a month on Haiku 5.5, $400 (₹37,680) on Sonnet 5.5, $800 (₹75,360) on Opus 5.5 and $1,600 (₹1,50,720) on Fable 5.2. The Batch API halves each figure. For routing tickets, Fable is the wrong tool.

3. A coding agent that reuses a large cached context

Assume 50 calls a day for 30 days, each reading 100,000 cached tokens of repository context plus 5,000 new input tokens and writing 3,000 output tokens. Because cache reads cost $0.20 per million on both Fable 5.2 and Opus 5.5, the month costs about $270 (₹25,434) on Fable, $150 (₹14,130) on Opus and $75 (₹7,065) on Sonnet, before cache writes. Fable’s premium comes almost entirely from output and fresh input.

Infographic of rupee costs for one long task, 1 lakh support tickets and a cached coding agent on Fable 5.2, Opus 5.5, Sonnet 5.5 and Haiku 5.5
Worked rupee examples at ₹94.20 to the dollar. · Infographic by 100xFounder

Prices exclude Indian taxes. How GST applies to a foreign API or subscription invoice depends on your business; confirm it with a chartered accountant.

Why it matters

Fable 5.2 cuts the price of Anthropic’s top tier by a fifth, but the bigger point for Indian teams is mixing models: the same IDs work across the Claude API, AWS, Google Cloud and Microsoft Foundry, and Anthropic itself says to start on Opus 5.5 and escalate. A team that sends routine traffic to Haiku or Sonnet and only the hardest steps to Fable can get frontier results at a rupee cost close to Sonnet’s. For a broader view of where AI budgets are going, see our analysis of OpenAI vs Anthropic enterprise spend, and for coding tools, our guide to the best AI for coding.

FAQ

When was Claude Fable 5.2 released?

Anthropic released Claude Fable 5.2 on 9 November 2026, together with Claude Mythos 5.2. The model’s API ID is claude-fable-5-2, and Anthropic says it won’t retire the model on its own platforms before 9 November 2027.

How much does Claude Fable 5.2 cost in India?

Through the API, Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, about ₹754 and ₹3,768 at ₹94.20 to the dollar, before tax. In the Claude apps, it is included for up to half the weekly limit on Max ($100 a month) and Ultra ($200), and billed through usage credits on Pro.

Is Claude Fable 5.2 better than Opus 5.5?

On Anthropic’s own benchmarks, Fable 5.2 beats Opus 5.5 on most tests, for example 49.1% vs 45.5% on FrontierCode 2.0. It also costs twice as much. Anthropic recommends starting on Opus 5.5 and moving to Fable only when Opus at high effort falls short.

Can I use Claude Fable 5.2 on AWS Bedrock or Google Cloud?

Yes. Fable 5.2 is on Amazon Bedrock (bedrock-runtime endpoint, global or US model ID), Claude Platform on AWS, Google Cloud’s Anthropic-managed and Google-managed Claude offerings, and Microsoft Foundry on Anthropic infrastructure, according to Anthropic’s documentation.

What is the difference between Claude Fable 5.2 and Mythos 5.2?

Anthropic says they are the same underlying model at the same price. Fable 5.2 has stronger cybersecurity and biology safeguards and is open to everyone; Mythos 5.2 is limited to organisations verified through Anthropic’s programmes, such as the Cyber Verification Program.

Does Claude Fable 5.2 support zero data retention?

No, not by default. Fable 5.2 requires 30-day data retention and is not available under zero data retention unless Anthropic expressly authorises it, according to its model overview page.

Interested in working on these models? See current openings on our Anthropic jobs page.

Sources

  1. Anthropic, “Introducing Claude Fable 5.2 and Claude Mythos 5.2”, 9 November 2026 (edited 11 November). anthropic.com
  2. Anthropic, “Claude Fable 5.2” model overview, Claude Platform docs, accessed 15 November 2026. platform.claude.com
  3. Anthropic, “What’s new in Claude Fable 5.2”, Claude Platform docs, accessed 15 November 2026. platform.claude.com
  4. Anthropic, “Models overview”, Claude Platform docs, accessed 15 November 2026. platform.claude.com
  5. Anthropic, “Pricing”, Claude Platform docs, accessed 15 November 2026. platform.claude.com
  6. Anthropic, “Choosing the right model”, Claude Platform docs, accessed 15 November 2026. platform.claude.com
  7. Anthropic, “Plans and pricing”, as shown in India, 15 November 2026. claude.com
  8. Claude Help Center, “Claude Fable 5.2 launch offers”, accessed 15 November 2026. support.claude.com
  9. Metropolitan Stock Exchange of India, “Historical Data for RBI Reference Rate”, 13 November 2026 rate. msei.in
Share this
Sources (9)
  1. Anthropic 9 Nov 2026Release with Mythos 5.2, same underlying model, safeguards and fallback to Opus 4.8/Opus 5, ~55% fewer blocks vs Fable 5.1, 30% faster output, $8/$40 and fast $16/$80, benchmark table (FrontierCode 2.0 49.1 vs 45.5, FrontierSWE v2 69.9 vs 62.3, AutomationBench 51.4 vs 42.5 and Gemini 4 Carbon 53.9, OSWorld 2.1 89.3 vs 87.2, HLE 74.2 vs 67.7), all platforms incl. AWS, Google Cloud, Azure, start with Opus 5.5 recommendation, Advisor role, launch week usage
  2. Anthropic 15 Nov 2026Released 9 Nov 2026, claude-fable-5-2, 1M context, 128K output, cache write/read prices, Bedrock ID, Jun 2026 cutoff, retirement not before 9 Nov 2027, 30-day retention and no ZDR unless authorised, forced tool use 400
  3. Anthropic 15 Nov 2026Four breaking changes, effort levels low-max default high, availability on Claude API, Claude Platform on AWS, Bedrock bedrock-runtime global/us IDs, Google Cloud both, Microsoft Foundry on Anthropic infrastructure, Fable 5.1 $10/$50, batch $4/$20
  4. Anthropic 15 Nov 2026Opus 5.5 $4/$20, Sonnet 5.5 $2/$10, Haiku 5.5 from $0.10/$0.50, latency labels, IDs, 1M context and 128K output for all four, start with Opus 5.5
  5. Anthropic 15 Nov 2026Full price table incl. Haiku 5.5 over/under 100K, cache multipliers, batch 50%, fast mode not with batch, long context standard pricing, inference_geo us 1.1x, regional endpoints +10%, tokenizer ~30% more tokens
  6. Anthropic 15 Nov 2026Model selection matrix and use cases; start with Opus 5.5; move to Fable 5.2 if xhigh/max evals fall short; effort tuning advice
  7. Anthropic 15 Nov 2026Shown in USD to Indian visitor before tax: Free no Fable; Pro $17/$20 Fable paid with usage credits; Max $100 and Ultra $200 Fable up to half weekly limit; Team standard $20/$25, premium $100/$125 50%; Enterprise $20/seat + API-rate usage, 20-seat minimum
  8. Anthropic (Claude Help Center) 15 Nov 2026Launch offers: Pro and standard Team 50% weekly on Fable until 16 Nov 11:59 pm PT plus one reset; Max/Ultra/premium 1-3 resets by tenure expiring 8 Mar 2027; billing before 9 Nov 10:00 am PT; resets don't raise limits; afterwards usage credits
  9. Metropolitan Stock Exchange of India 13 Nov 2026RBI reference rate USD/INR 94.1997 on 13 Nov 2026

Related reading

All stories →
Was this page useful? Be the first to rate it

Watch our reels

Startup news, founder stories and AI tools in under a minute.

Follow @100x.founder on Instagram
Everyone says ships can hide at sea by switching off trackers. This satellite is built to find them. Everyone says your mind is the one private place. AI can already redraw what you saw. Everyone says money can't buy genes. A startup says it can, for a fee. Everyone says a service that saves lives should stay free. Blinkit Ambulance is about to charge. Everyone says gold smuggling is a crime problem. This says it is a tax problem. Everyone says quit when the numbers say quit. This says a dream that survives is a win.