# Claude Opus 5.5 pricing: $4 input, $20 output per million tokens

> Opus 5.5 costs 20% less per token than Opus 5, and Anthropic says typical jobs cost 40% less to run. Here are the numbers as of October 28, 2026.

URL: https://claudegpt.si/news/claude-opus-5-5-pricing  
Published: 2026-10-28T10:56:33.568Z  
Author: ClaudeGPT Si Newsroom  
Section: AI Models  
Topics: Claude, Anthropic, Claude Opus 5.5, AI pricing  
Publisher: ClaudeGPT Si

## Key takeaways

- As of October 28, 2026, Claude Opus 5.5 costs $4 per million input tokens and $20 per million output tokens, 20% below Opus 5.
- Cache reads cost $0.20 per million tokens, down from $0.50, and the Batch API halves input and output prices.
- Anthropic says Opus 5.5 costs 40% less to run than Opus 5 on typical workloads at default settings, a claim from its own tests.

Claude Opus 5.5 pricing is $4 per million input tokens and $20 per million output tokens, as of October 28, 2026. That is 20% below Claude Opus 5, which costs $5 and $25, according to Anthropic's [pricing page](https://platform.claude.com/docs/en/about-claude/pricing). So yes, Opus 5.5 is cheaper than Opus 5, and the gap is bigger for cached input.

Anthropic [released Opus 5.5 on September 22](https://www.anthropic.com/news/claude-opus-5-5) and says it costs 40% less to run than Opus 5 on typical workloads at default settings. That is a claim about the cost of a whole job, from Anthropic's own tests. The 20% price cut is the part you can check on the price list. For context on the rest of the lineup, see our [recap of the Claude 5.5 releases](https://claudegpt.si/news/claude-news-5-5-family-recap).

## Claude Opus 5.5 token cost per million tokens

These are standard API prices in US dollars per million tokens (MTok), as of October 28, 2026, from Anthropic's [pricing docs](https://platform.claude.com/docs/en/about-claude/pricing).

| | Opus 5.5 | Opus 5 |
| --- | --- | --- |
| Input | $4 | $5 |
| Output | $20 | $25 |
| Cache write, 5 minutes | $5 | $6.25 |
| Cache write, 1 hour | $8 | $10 |
| Cache read | $0.20 | $0.50 |
| Batch input | $2 | $2.50 |
| Batch output | $10 | $12.50 |

Prompt caching lets you store part of a prompt and pay less to reuse it. Cache reads on Opus 5.5 cost 5% of the input price, where most earlier models charge 10%. That is why cache reads fall 60%, from $0.50 to $0.20. A workload that reuses a long system prompt or a large codebase on every call benefits the most.

The Batch API, which runs requests asynchronously, takes 50% off both input and output.

## Is Claude Opus 5.5 cheaper than Opus 5?

On the price list, yes, by 20% on input and output. Here is a simple example. A request with 50,000 input tokens and 15,000 output tokens costs $0.50 on Opus 5.5 ($0.20 for input, $0.30 for output). The same request costs $0.625 on Opus 5. That is our arithmetic from the published prices, not an Anthropic figure.

Your real bill depends on how many tokens the model uses, and two settings changed. Opus 5.5 always runs with adaptive thinking, which means the model decides how much to think, and you cannot turn it off. Thinking tokens are billed as output. The default effort level is `medium`, where Opus 5's was `high`, according to the [migration guide](https://platform.claude.com/docs/en/models/opus-5-5/migration-guide). Anthropic's 40% figure refers to default settings. If you raise effort to `xhigh` or `max`, expect to spend more per request. Anthropic advises re-measuring cost and latency at the effort level you choose.

## Fast mode, US-only inference and long context

- **Fast mode.** Opus 5.5 offers up to 2.5 times faster output at $8 per million input tokens and $40 per million output tokens, as of October 28, 2026. That is twice the standard price. It is not available with the Batch API.
- **US-only inference.** Setting `inference_geo` to `us` applies a 1.1x multiplier to all token prices.
- **Long context.** Opus 5.5 has a 1M-token context window, and the docs say it is billed at the standard rate across the whole window. A 900,000-token request costs the same per token as a 9,000-token one.

## How Opus 5.5 compares with other Claude models

| Model | Input | Output | API ID |
| --- | --- | --- | --- |
| Claude Fable 5.1 | $10 | $50 | `claude-fable-5-1` |
| Claude Opus 5.5 | $4 | $20 | `claude-opus-5-5` |
| Claude Sonnet 5.5 | $2 | $10 | `claude-sonnet-5-5` |
| Claude Haiku 5.5 | from $0.10 | from $0.50 | `claude-haiku-5-5` |

Prices per million tokens as of October 28, 2026, from Anthropic's [models overview](https://platform.claude.com/docs/en/about-claude/models/overview). Haiku 5.5 is priced higher for prompts over 100,000 tokens ($0.50 and $2.50).

Anthropic's docs suggest starting with Opus 5.5 for most workloads, and moving to Fable 5.1 when your own evaluations at higher effort still fall short. Opus 5.5 costs 40% of Fable 5.1 per token. For the cheaper tier, see our piece on [what Sonnet 5.5 costs](https://claudegpt.si/news/claude-sonnet-5-5-cost).

The model ID is `claude-opus-5-5`. It is available on the Claude Platform, Amazon Web Services, Google Cloud and Microsoft Azure. Partner platforms can set their own prices and regional premiums, so check the pricing page of the platform you use. Anthropic says the five-hour usage limits on Pro, Max, Team and seat-based Enterprise plans are going up with this release, but it gave no per-plan numbers in the announcement.

## What to check before you switch

Prices on Anthropic's page have changed between model generations, so check the pricing page again before you commit to a budget. Run your own evaluation at `medium` effort, compare the cost per finished task rather than the price per token, and note that requests that set `thinking` to disabled return an error on Opus 5.5 and need a code change.

## Frequently asked questions

### How much does Claude Opus 5.5 cost?

As of October 28, 2026, the API charges $4 per million input tokens and $20 per million output tokens. Batch requests cost $2 and $10, and fast mode costs $8 and $40.

### Is Claude Opus 5.5 cheaper than Opus 5?

Yes. Opus 5 costs $5 and $25 per million tokens, so Opus 5.5 is 20% cheaper per token. Anthropic also says it costs 40% less to run on typical workloads at default settings.

### What is the Claude Opus 5.5 token cost for cached prompts?

Cache reads cost $0.20 per million tokens and five-minute cache writes cost $5, as of October 28, 2026. Opus 5 charges $0.50 and $6.25.

### Does Claude Opus 5.5 charge extra for long context?

No. Anthropic's docs say the full 1M-token window is billed at the standard per-token rate for Opus 5.5.

## Sources

- [Claude Opus 5.5 announcement](https://www.anthropic.com/news/claude-opus-5-5), Anthropic
- [Pricing](https://platform.claude.com/docs/en/about-claude/pricing), Anthropic
- [Claude pricing](https://claude.com/pricing), Anthropic
- [Models overview](https://platform.claude.com/docs/en/about-claude/models/overview), Anthropic
- [Claude Opus 5.5 migration guide](https://platform.claude.com/docs/en/models/opus-5-5/migration-guide), Anthropic