# Claude Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 4.5 compared

URL: https://infro.io/blog/anthropic-model-lineup-2026
Author: INFRO Editorial (https://infro.io/blog/editorial)
Publisher: INFRO
Published: 2026-09-29
Last modified: 2026-09-29
Last reviewed: 2026-09-29 (next review due 2026-10-29)
Topic: Evaluation and models
Format: Comparison

> Anthropic's current models are Claude Fable 5.1 ($10/$50 per million input/output tokens), Opus 5.5 ($4/$20), Sonnet 5.5 ($2/$10) and Haiku 4.5 ($1/$5). The first three have a 1M-token window and 128K output; Haiku 4.5 has 200K and 64K, an older knowledge cutoff and a retirement commitment of not sooner than October 15, 2026. Pick per task by evaluating from the lower-cost model upward, and plan a Haiku 4.5 migration path now.

Key takeaways:
- Standard rates run from $1/$5 on Haiku 4.5 to $10/$50 on Fable 5.1 per million input and output tokens.
- Cache reads are not a flat 10%: 0.025x input on Fable 5.1 and 0.05x on Opus 5.5 put both at or near Sonnet 5.5's cache-read rate.
- Haiku 4.5 is the outlier: 200K context, 64K output, extended rather than adaptive thinking, and the earliest retirement commitment.
- Anthropic suggests Opus 5.5 as the starting point for most workloads; that is a default to test, not a decision.
- Route by task tier and keep a second Claude model behind each route, so a retirement or rate change is a configuration edit.

Anthropic's current lineup has four models: Claude Fable 5.1, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 4.5. All four take text and image input, produce text, and support tool use and vision[1]. Beyond that they differ in almost every published dimension: list rate, cache-read rate, context window, output limit, thinking mode, knowledge cutoff and how long Anthropic commits to keep them.

This article sets out those published facts, read on 2026-09-29, and a method for deciding which model runs which task. It contains no quality ranking; Anthropic's own descriptions are quoted as descriptions, not as verdicts.

## How Anthropic describes each model

Anthropic's overview gives each model a one-line purpose and a comparative latency[1]:

- **Claude Fable 5.1** — for demanding reasoning and long-horizon agentic work. Latency: slower.
- **Claude Opus 5.5** — for long-running agentic coding and knowledge work. Latency: moderate.
- **Claude Sonnet 5.5** — described as the best combination of speed and intelligence. Latency: fast.
- **Claude Haiku 4.5** — the fastest model, with near-frontier intelligence by Anthropic's account. Latency: fastest.

Anthropic also gives a default: start with Opus 5.5 for most workloads, and use Fable 5.1 when evaluations on Opus 5.5 at higher effort still fall short[1]. Treat that as the vendor's suggested starting point for your own evaluation.

## List rates

[Diagram: Anthropic list rates per million tokens, from its pricing page on 2026-09-29.]

Per million tokens, Fable 5.1 lists $10 input and $50 output, Opus 5.5 $4 and $20, Sonnet 5.5 $2 and $10, and Haiku 4.5 $1 and $5[2]. The catalog entries INFRO estimates against are on the [Claude Fable 5.1](https://infro.io/models/anthropic/claude-fable-5.1), [Claude Opus 5.5](https://infro.io/models/anthropic/claude-opus-5.5), [Claude Sonnet 5.5](https://infro.io/models/anthropic/claude-sonnet-5.5) and [Claude Haiku 4.5](https://infro.io/models/anthropic/claude-haiku-4.5) model pages.

The Batch API halves both rates for all four: $5/$25, $2/$10, $1/$5 and $0.50/$2.50[2]. Opus 5.5 also has a fast mode in research preview at $8 input and $40 output, on the Claude API only[2].

## Cache reads are not uniform

Most Claude models bill a cache hit at 0.1x the input rate. Two current models do not: Fable 5.1 bills cache hits at 0.025x input ($0.25 per million) and Opus 5.5 at 0.05x ($0.20 per million)[2]. Sonnet 5.5 is at the standard 0.1x, also $0.20, and Haiku 4.5 at $0.10[2].

So for a workload dominated by cache reads — an agent resending a large, stable context — the input side of the bill is similar across Fable 5.1, Opus 5.5 and Sonnet 5.5. The difference then sits mostly in uncached input and output.

> **Example: A cache-heavy agent step**
> Illustrative arithmetic on published rates. Each step sends 80,000 tokens, of which 76,000 are cache hits, and generates 2,000 tokens.
> 
> Fable 5.1: 4,000 × $10 + 76,000 × $0.25 + 2,000 × $50, per million = $0.04 + $0.019 + $0.10 = $0.159.
> 
> Opus 5.5: $0.016 + $0.0152 + $0.04 = $0.0712.
> 
> Sonnet 5.5: $0.008 + $0.0152 + $0.02 = $0.0432.
> 
> Haiku 4.5 cannot take this step at all if the full context exceeds its 200K window; here it fits: $0.004 + $0.0076 + $0.01 = $0.0216. These figures assume identical output length and success, which you have to measure.

Cache writes follow Anthropic's multipliers: 1.25x input for a 5-minute write and 2x for a 1-hour write, so $5 and $8 per million on Opus 5.5, for example[2].

## Limits, thinking and knowledge

[Diagram: Published limits and controls for Anthropic's four current models, read 2026-09-29.]

- **Context and output.** Fable 5.1, Opus 5.5 and Sonnet 5.5 have a 1M-token window and 128K output on the synchronous Messages API; Haiku 4.5 has 200K and 64K[1]. On the Message Batches API, Opus 5.5 and Sonnet 5.5 support up to 300K output tokens with a beta header[1].
- **Long prompts.** Claude 4.6 and later models include the full 1M window at standard pricing[2]; there is no long-prompt surcharge on the three 1M models.
- **Thinking.** Fable 5.1 and Opus 5.5 use adaptive thinking that is always on; Sonnet 5.5 uses adaptive thinking; Haiku 4.5 uses extended thinking. Default effort is `high` on Fable 5.1 and Sonnet 5.5, `medium` on Opus 5.5, and the effort parameter is not supported on Haiku 4.5[1].
- **Knowledge.** Reliable knowledge runs to June 2026 on the three newest models and to February 2025 on Haiku 4.5[1].
- **Structured outputs.** Generally available on all four[3].
- **Tokenizer.** Claude 4.7 and later models use a tokenizer that produces approximately 30% more tokens for the same text[2]. When comparing Haiku 4.5 against the newer models, compare measured tokens per task, not rates alone.

## Overheads that differ by model

Two published overheads change the token count of the same request from one Claude model to the next, which matters most for small, frequent calls.

**Tool definitions.** When a request includes tools, Anthropic adds a tool-use system prompt whose size depends on the model: 286 tokens on Opus 5.5 and Sonnet 5.5, and 496 tokens on Haiku 4.5 with `tool_choice` set to `auto` or `none`, or 588 with `any` or `tool`[2]. That is before your own tool names, descriptions and schemas, which are billed as input too. For a classification call of a few hundred tokens with one tool attached, the fixed overhead can be a large share of the input.

**Tokenizer.** The newer tokenizer on Claude 4.7 and later models produces approximately 30% more tokens for the same text, with the exact increase depending on content[2]. Haiku 4.5 predates it. A rate comparison between Haiku 4.5 and Sonnet 5.5 therefore understates the difference in tokens counted for the same prompt; measure tokens per task on your own inputs rather than converting by rule of thumb.

## The same models on other clouds

Anthropic publishes model ids for Amazon Bedrock, Google Cloud, Microsoft Foundry and Claude Platform on AWS for all four models — for example `anthropic.claude-opus-5-5` on Bedrock and `claude-haiku-4-5@20251001` on Google Cloud[1]. The commercial terms differ by route:

- **Partner-operated clouds** (Bedrock, Google Cloud) invoice you themselves and set their own prices and lifecycle dates; their regional and multi-region endpoints carry a 10% premium over global ones[2].
- **Anthropic-operated platforms** billed through your cloud account rate usage at the standard per-model rates[2].
- **US-only inference** on the Claude API with `inference_geo: "us"` applies a 1.1x multiplier to every token category[2].
- **Fast mode** for Opus 5.5 is available on the Claude API only[2].

If your organization already buys Claude through a cloud commitment, the list rate on this page may not be what you pay. Compare against the rate your contract sets.

## Retirement commitments

[Diagram: Anthropic's commitments for its own platforms: each model will not be retired sooner than the date shown. Partner clouds set their own dates.]

Anthropic commits not to retire Fable 5.1 sooner than September 1, 2027, Opus 5.5 sooner than September 22, 2027, Sonnet 5.5 sooner than September 28, 2027, and Haiku 4.5 sooner than October 15, 2026[1]. The Haiku 4.5 date is a floor, not an announced retirement, but it is two and a half weeks from the date of this article. Any workload on Haiku 4.5 should have a tested successor ready.

## A method for choosing

1. **Group work by task type**, not by application: classification, extraction, support drafting, research agents and code changes each get their own decision.
2. **Evaluate upward.** Start with the lowest-cost model that meets the hard constraints (window, output limit, latency), then move up only when it fails your check. Anthropic's own guidance on its pricing page follows the same shape: Haiku for simple tasks, Sonnet for most production workloads, Opus for the most complex reasoning[2].
3. **Test effort as well as model.** Opus 5.5 at higher effort and Fable 5.1 at its default are different operating points; Anthropic suggests exactly that comparison[1].
4. **Measure cost per successful task** at the effort you will ship, with caching on.
5. **Record the second choice.** For every task tier, keep a tested alternative; it is your failover target and your migration path.

## Routing Claude models with INFRO

Connect your Anthropic account as a [provider connection](https://infro.io/docs/providers); your provider bills you at your rates. INFRO's Bedrock connection speaks Chat Completions, which Bedrock does not offer for Claude, so Claude bought through AWS is not routable through INFRO yet. Anthropic notes that partner clouds price Claude independently[2]; if you have a negotiated rate, set it under Prices ([billing](https://infro.io/docs/org/billing)).

Then put each task tier behind a [route](https://infro.io/docs/routes):

| Route slug | Targets, in order |
| --- | --- |
| `support-drafts` | `anthropic/claude-sonnet-5.5` on Anthropic, then the same model on Bedrock |
| `research-agent` | `anthropic/claude-opus-5.5`, then `anthropic/claude-fable-5.1` |
| `ticket-tags` | `anthropic/claude-haiku-4.5`, then `anthropic/claude-sonnet-5.5` |

Table: An example route plan across Anthropic models and connections.

The first route fails over to the same model on another account, which changes infrastructure and not behaviour. The other two fail over to a different model, which changes behaviour, so list only a model you have evaluated on that task ([failover](https://infro.io/docs/features/fallbacks)). When Haiku 4.5 is retired, `ticket-tags` becomes a one-line edit: move its tested successor to first place.

Use a [model allowlist](https://infro.io/docs/org/model-allowlists) to keep `anthropic/claude-fable-5.1` to the projects approved for it; a route target that fails the allowlist is skipped. For Anthropic prompt caching, put `cache_control` on the text part that ends your stable prefix; INFRO passes it through, and `usage.prompt_tokens_details.cached_tokens` shows whether the cache was read ([prompt caching](https://infro.io/docs/features/prompt-caching)). `GET /v1/usage?group_by=route` then gives estimated provider cost, error rate and latency per route, so the cost of each tier decision is visible month to month ([usage analytics](https://infro.io/docs/observability/analytics)).

## Questions

### What does Claude fast mode cost on Opus 5.5?

Anthropic lists fast mode, in research preview, for Claude Opus 5.5 at $8 per million input tokens and $40 per million output tokens on the Claude API only; it is not available with the Batch API or on partner clouds.

### Do newer Claude models use more tokens for the same text?

Anthropic's pricing page says Claude 4.7 and later models use a newer tokenizer that produces approximately 30% more tokens for the same text, depending on content. Compare models on measured cost per task rather than on rate alone when one side uses the older tokenizer.

## Sources

1. Models overview — Anthropic. https://platform.claude.com/docs/en/about-claude/models/overview (accessed 2026-09-29)
2. Pricing — Anthropic. https://platform.claude.com/docs/en/about-claude/pricing (accessed 2026-09-29)
3. Structured outputs — Anthropic. https://platform.claude.com/docs/en/build-with-claude/structured-outputs (accessed 2026-09-29)