All Claude AI models available in 2026: Opus 5.5, Fable 5.2, Sonnet 5.5 and Haiku 5.5
Updated: 5 days ago

Updated: November 15, 2026
Anthropic's current Claude lineup has four general models: Claude Fable 5.2, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5, plus Claude Mythos 5.2, which shares Fable 5.2's specifications and pricing but is reserved for organizations verified through Anthropic's verification programs. All five have a 1M-token context window and 128K tokens of maximum output, and on the Claude API they cost from $0.10 per million input tokens (Haiku 5.5 on short prompts) up to $8 input and $40 output (Fable 5.2 and Mythos 5.2).
In the Claude apps, the Free plan includes Sonnet and Haiku, while Pro, Max, Ultra and Team include Opus, Sonnet and Haiku, and Anthropic says Opus 5.5 is now the default model on Pro, Max, Ultra and Team. Fable is available on every paid plan, but on Pro and on standard Team seats it runs on usage credits. Below are the full model table with API IDs and prices, the plan-by-plan access rules, the older models that still answer API calls, the retirement calendar and a monthly cost example for each current model.
........
Model | Released | API ID | Input / output per 1M tokens | Context / max output | Access |
Claude Fable 5.2 | Nov 9, 2026 | claude-fable-5-2 | $8 / $40 | 1M / 128K | API, all paid app plans |
Claude Mythos 5.2 | Nov 9, 2026 | claude-mythos-5-2 | $8 / $40 | 1M / 128K | Verified organizations only |
Claude Opus 5.5 | Sep 22, 2026 | claude-opus-5-5 | $4 / $20 | 1M / 128K | API, Pro and above |
Claude Sonnet 5.5 | Sep 28, 2026 | claude-sonnet-5-5 | $2 / $10 | 1M / 128K | API, all app plans |
Claude Haiku 5.5 | Oct 7, 2026 | claude-haiku-5-5 | from $0.10 / $0.50 | 1M / 128K | API, all app plans |
........
··········
THE FOUR CURRENT CLAUDE MODELS AND WHAT EACH ONE IS FOR.
Anthropic's model overview tells developers who are unsure to start with Claude Opus 5.5 for most workloads and to move to Claude Fable 5.2 for demanding reasoning and long-horizon agentic work, or when evaluations on Opus 5.5 at a higher effort setting still fall short. Opus 5.5 launched on September 22, 2026 at $4 per million input tokens and $20 per million output tokens, half the price of Fable 5.2, with always-on adaptive thinking and a default effort of medium. Anthropic describes it as the model for long-running agentic coding and knowledge work, and in a November 12 article it recommends Opus 5.5 as the daily model for research, analysis and finished documents, decks and spreadsheets.
Claude Fable 5.2 arrived on November 9, 2026 together with Claude Mythos 5.2. It costs $8 input and $40 output per million tokens, 20% less than Fable 5.1 at $10 and $50, runs adaptive thinking that cannot be switched off, and uses a default effort of high. Mythos 5.2 offers the same capabilities only to organizations admitted to programs such as the Cyber Verification Program, so most API customers and every consumer plan see Fable, not Mythos. Our report on the Claude Fable 5.2 and Mythos 5.2 launch covers the launch offers and the price cut in detail.
Claude Sonnet 5.5, released on September 28, 2026, is the model Anthropic calls the best combination of speed and intelligence, at $2 input and $10 output. In the app guidance Anthropic positions it as faster and cheaper than Opus 5.5 for well-scoped tasks with a clear brief, and suggests switching to Opus 5.5 when a task needs Sonnet pushed to its highest effort levels. Claude Haiku 5.5, released on October 7, 2026, is the fastest model of the lineup and targets high-volume, latency-sensitive work such as classification, extraction, routing and subagent tasks.
All four current models accept text and image input, produce text output, support tool use and vision, and share a reliable knowledge cutoff of June 2026. Every one of them now has the same 1M-token context window and 128K-token output ceiling on the synchronous Messages API, and on the Message Batches API Opus 5.5, Sonnet 5.5 and Haiku 5.5 can return up to 300K output tokens with Anthropic's output-300k-2026-03-24 beta header.
··········
CLAUDE MODEL ACCESS ON FREE, PRO, MAX, ULTRA, TEAM AND ENTERPRISE PLANS.
The official pricing page lists model families per plan without version numbers. The Free plan includes Sonnet and Haiku and excludes Opus and Fable. Pro ($20 a month, or $17 a month billed annually) adds Opus, and lists Fable as paid with usage credits. Max ($100 a month, five times Pro usage) and Ultra ($200 a month) include Opus, Sonnet and Haiku and let Fable use up to half of the weekly limit at no extra cost.
........
Plan | Price | Models included | Fable access |
Free | $0 | Sonnet, Haiku | Not available |
Pro | $20/month or $17/month annual | Opus, Sonnet, Haiku | Usage credits only |
Max | $100/month | Opus, Sonnet, Haiku | Up to 50% of weekly limit |
Ultra | $200/month | Opus, Sonnet, Haiku | Up to 50% of weekly limit |
Team standard seat | $25/month or $20/month annual | Opus, Sonnet, Haiku | Usage credits only |
Team premium seat | $125/month or $100/month annual | Opus, Sonnet, Haiku | Up to 50% of weekly limit |
Enterprise | $20/seat/month annual plus usage at API rates | Fable, Opus, Sonnet, Haiku | Depends on seat type and admin settings |
........
Anthropic's help article on Fable models says Fable 5.2, 5.1 and 5 are available on all paid plans and not on Free. The 50% figure on Max, Ultra and Team premium seats is a cap on the share of the weekly limit Fable may consume, and Fable draws down that limit faster than other models, so heavy Fable use leaves less room for Opus and Sonnet in the same week. For a limited time after the November 9 launch, eligible paid subscribers also receive extra Fable 5.2 usage; the article does not give an end date.
Enterprise organizations get Fable 5.2 switched off until an administrator enables it. Seven days after launch, Anthropic adds it to roles that already have any Fable model enabled and moves Fable defaults to 5.2 unless the administrator opts out first. Claude Code needs version 2.1.321 or later to use Fable 5.2. The exact weekly limits for each plan are covered in our guide to Claude plans and usage limits.
··········
CLAUDE API PRICES FOR CURRENT AND OLDER MODELS.
Anthropic's pricing page lists every model still billable on the Claude API. Batch requests cost 50% less, US-only inference costs 1.1 times the standard price, and fast mode on Opus 5.5 and Fable 5.2 costs twice the standard price. Cache read prices differ by model, at 2.5% of base input on Fable 5.2 and 5% on Opus 5.5 and Sonnet 5.5. The cache write column below is the five-minute cache.
........
Model | Input | Output | Cache write (5 min) | Cache read |
Fable 5.2 | $8 | $40 | $10 | $0.20 |
Opus 5.5 | $4 | $20 | $5 | $0.20 |
Sonnet 5.5 | $2 | $10 | $2.50 | $0.10 |
Haiku 5.5 (prompts up to 100K tokens) | $0.10 | $0.50 | $0.125 | $0.01 |
Haiku 5.5 (prompts over 100K tokens) | $0.50 | $2.50 | $0.625 | $0.05 |
Fable 5.1 | $10 | $50 | $12.50 | $0.25 |
Fable 5 | $10 | $50 | $12.50 | $1 |
Opus 5, 4.8, 4.7, 4.6, 4.5 | $5 | $25 | $6.25 | $0.50 |
Sonnet 5 | $2 | $10 | $2.50 | $0.20 |
Sonnet 4.6 and 4.5 | $3 | $15 | $3.75 | $0.30 |
Haiku 4.5 | $1 | $5 | $1.25 | $0.10 |
........
Prices are in USD per million tokens as listed on Anthropic's official pricing page on November 15, 2026. Haiku 5.5 is the only current model with a price break by prompt size: up to 100,000 tokens it costs $0.10 input and $0.50 output, and above that threshold every rate rises fivefold. Haiku 5.5 also uses the newer tokenizer introduced with Claude Opus 4.7, which Anthropic says counts the same text as roughly 30% more tokens than Haiku 4.5 does, a factor to include when comparing bills across generations.
··········
[[ADPLUS_300x250_1]]
CLAUDE MODEL IDS ON AMAZON BEDROCK, GOOGLE CLOUD AND MICROSOFT FOUNDRY.
From the 4.6 generation on, every Claude model ID is a pinned snapshot without a date, so claude-opus-5-5 will not silently change to a newer model. The same IDs work on Claude Platform on AWS, on Claude Models on Google Cloud (both the Anthropic-managed and the Google-managed versions) and on Microsoft Foundry, where the deployment name defaults to the Claude API ID. Amazon Bedrock uses its own prefixes.
........
Model | Claude API | Amazon Bedrock | Google Cloud | Microsoft Foundry |
Fable 5.2 | claude-fable-5-2 | global.anthropic.claude-fable-5-2 | claude-fable-5-2 | claude-fable-5-2 |
Opus 5.5 | claude-opus-5-5 | anthropic.claude-opus-5-5 | claude-opus-5-5 | claude-opus-5-5 |
Sonnet 5.5 | claude-sonnet-5-5 | anthropic.claude-sonnet-5-5 | claude-sonnet-5-5 | claude-sonnet-5-5 |
Haiku 5.5 | claude-haiku-5-5 | anthropic.claude-haiku-5-5 | claude-haiku-5-5 | claude-haiku-5-5 |
........
Anthropic's retirement dates apply to the platforms it operates: the Claude API, Claude Platform on AWS, the Anthropic-managed Google Cloud offering and Microsoft Foundry. Amazon Bedrock and the Google-managed Google Cloud offering set their own lifecycle dates, which is why some models already retired on the Claude API, such as Opus 4.1 and Sonnet 4, still run on Bedrock or Google Cloud. Developers can read each model's token limits and capabilities programmatically through the Models API, which returns max_input_tokens, max_tokens and a capabilities object for every available model.
··········
OLDER CLAUDE MODELS: LEGACY, DEPRECATED AND RETIRED.
Many earlier models still answer API calls. Anthropic lists Fable 5.1, Fable 5, Opus 5, Opus 4.8, Opus 4.7, Opus 4.6, Opus 4.5, Sonnet 5, Sonnet 4.6 and Haiku 4.5 as active, but recommends the latest models wherever possible. Claude Sonnet 4.5 is deprecated and will stop working on the Claude API on November 30, 2026, with Sonnet 5.5 as the recommended replacement. Claude Opus 4.5 and Claude Haiku 4.5 carry the earliest possible retirement dates among active models, not sooner than November 24, 2026 and October 15, 2026 respectively, although neither has been deprecated yet.
........
Model | Status on the Claude API | Date | Recommended replacement |
Claude Sonnet 4.5 | Deprecated | Retires Nov 30, 2026 | claude-sonnet-5-5 |
Claude Opus 4.1 | Retired | Aug 5, 2026 | claude-opus-4-8 |
Claude Opus 4 | Retired | Jun 15, 2026 | claude-opus-4-8 |
Claude Sonnet 4 | Retired | Jun 15, 2026 | claude-sonnet-5-5 |
Claude Haiku 3 | Retired | Apr 20, 2026 | claude-haiku-4-5-20251001 |
Claude Sonnet 3.7 | Retired | Feb 19, 2026 | claude-sonnet-5-5 |
Claude Haiku 3.5 | Retired | Feb 19, 2026 | claude-haiku-4-5-20251001 |
Claude Opus 3 | Retired | Jan 5, 2026 | claude-opus-4-8 |
........
Anthropic commits to at least 60 days' notice before retiring a publicly released model and notifies customers with active deployments by email. Requests to a retired model fail, so teams can audit their usage from the Usage page of the Claude Console, where an export breaks down usage by API key and model. Claude Mythos Preview, the predecessor of the Mythos line, has been deprecated since June 9, 2026 with a retirement date still to be announced.
API version 2026-10-21 also removed the temperature, top_p and top_k parameters, which now return a 400 error. Earlier API versions return the same error when those parameters are set to non-default values on Claude 4.7 and later models, so code written for Sonnet 4 or Opus 4.1 usually needs these fields deleted when it moves to the 5.5 models.
··········
WHAT A CHATBOT WORKLOAD COSTS ON EACH CLAUDE MODEL.
To compare the price levels, consider a customer-support assistant that handles 10,000 requests a day for 30 days, with 3,000 input tokens and 500 output tokens per request and no prompt caching. That is 900 million input tokens and 150 million output tokens a month. The arithmetic below uses the official list prices; the Batch column applies the 50% discount, which suits offline jobs and does not fit live chat.
........
Model | Monthly cost, standard | Monthly cost, Batch API |
Fable 5.2 | $13,200 | $6,600 |
Opus 5.5 | $6,600 | $3,300 |
Sonnet 5.5 | $3,300 | $1,650 |
Haiku 5.5 (prompts up to 100K) | $165 | $82.50 |
Sonnet 4.5 (retiring Nov 30) | $4,950 | $2,475 |
Haiku 4.5 | $1,650 | $825 |
........
On these assumptions Haiku 5.5 costs one tenth of Haiku 4.5 per token, and a team still on Sonnet 4.5 pays a third less by moving to Sonnet 5.5 before the November 30 retirement. The tokenizer difference narrows the Haiku gap in practice, because the same conversation produces about 30% more tokens on Haiku 5.5, but the newer model remains far cheaper at list prices. Prompt caching changes the totals further for assistants that reuse a long system prompt, since a cache read on Opus 5.5 or Fable 5.2 costs $0.20 per million tokens instead of $4 or $8.
··········
HOW TO CHOOSE AND CHECK A CLAUDE MODEL.
In the Claude apps the model selector sits next to the message box, and Anthropic's own advice is to raise the effort setting before moving up a model. Opus 5.5 is the starting point for most paid users, Haiku 5.5 suits quick lookups and short, narrowly scoped tasks, and Fable 5.2 fits projects whose scope and complexity exceed what Opus 5.5 handles at higher effort. Free users can choose between the Sonnet and Haiku models the plan includes.
Developers can confirm availability by calling the Models API with their key, which lists only the models their organization can use, including any limited-access Mythos model. All current models offer the same 1M-token window, and our guide to the Claude context window and token limits explains how that window behaves in long conversations. For production use, Anthropic recommends testing a replacement well before the retirement date of the model in use and pinning the dateless ID of the chosen version.
··········
THE CLAUDE MODEL LINEUP IN NOVEMBER 2026.
The current Claude family is simpler than a year ago. Four models share one 1M-token window and one 128K output limit, and they differ mainly by price, speed and depth of reasoning: Haiku 5.5 from $0.10 input, Sonnet 5.5 at $2, Opus 5.5 at $4 and Fable 5.2 at $8. Mythos 5.2 matches Fable 5.2 for verified organizations only. In the apps, Free covers Sonnet and Haiku, paid plans default to Opus 5.5, and Fable is a usage-credit model on Pro and standard Team seats.
The open points are dates. Anthropic has not said when the Fable 5.2 launch offer ends, the pricing page does not name the exact Sonnet and Haiku versions served on the Free plan, and Opus 4.5 and Haiku 4.5 have passed or are close to their earliest retirement dates without a deprecation notice. Sonnet 4.5 is the one firm deadline, with API calls to it ending on November 30, 2026.
··········
FOLLOW US FOR MORE.
·····
DATA STUDIOS
·····
datastudios.org




