Anthropic: Claude Opus 5.5
Claude Opus 5.5 is a reasoning model from Anthropic released on Sep 18, 2026. It reads text, images and files and writes text, with a 1M token context window and up to 128K tokens per reply.
Show moreShow less
It costs $4 per million input tokens and $20 per million output tokens. Cached input bills at $0.20, 95% less than fresh input.
It supports tool calling, structured JSON output and prompt caching.
28 providers serve it, including Amazon Bedrock, Azure and Vertex.
On the Artificial Analysis Intelligence Index it scores 57.6, which places it first among the 43 models I track with independent scores.
- Modalities
- In / out price
- $4 / $20 per 1M
- Context
- 1M
- Max output
- 128K
- Released
- Sep 18, 2026
- Knowledge cutoff
- Jun 2026
- Tool calling
- Structured output
- Reasoning
- Prompt caching
- Open weights
- Free to use
Claude Opus 5.5 pricing
USD, as listed on Oct 15, 2026. Most bills are input heavy, so the cached input rate matters as much as the headline price.
| Input | $4 / 1M tokens | |
|---|---|---|
| Output | $20 / 1M tokens | |
| Cached input (read) | $0.20 / 1M tokens | 95% off input |
| Cache write | $5 / 1M tokens |
Claude Opus 5.5 providers
28 providers serve it, including Amazon Bedrock, Azure and Vertex. Each host sets its own price and limits, so the cheapest one is not always the one with the longest context.
| Provider | Input /M | Output /M | Cache read /M | Context | Max output |
|---|---|---|---|---|---|
| Anthropic First party | $4 | $20 | $0.20 | 1M | 128K |
| Amazon Bedrock | $4 | $20 | $0.20 | 1M | 128K |
| Azure | $4 | $20 | $0.20 | 1M | 128K |
| Vertex | $4 | $20 | $0.20 | 1M | 128K |
| Vertex (Anthropic) | $4 | $20 | $0.20 | 1M | 128K |
| Vercel AI Gateway | $8 | $40 | $0.40 | 1M | 128K |
| 302.AI | $4 | $20 | $0.20 | 1M | 128K |
| AIHubMix | $4 | $20 | $0.20 | 1M | 128K |
| Azure Cognitive Services | $4 | $20 | $0.20 | 1M | 128K |
| Cloudflare AI Gateway | $4 | $20 | $0.20 | 1M | 128K |
| CrossModel | $4 | $20 | $0.20 | 1M | 128K |
| DevPass (LLM Gateway) | $4 | $20 | $0.20 | 1M | 128K |
Showing 12 of 28 providers: the lab first, then the major clouds, then the cheapest resellers.
What Claude Opus 5.5 costs per month
Pick a workload close to yours or type your own volumes.
Independent benchmarks
Measured by third parties, not reported by Anthropic. Each links to its source and shows the date I checked it.
A composite of 10 independent evaluations (agents 30%, coding 20%, scientific reasoning 20%, general 30%), run by Artificial Analysis on each model, version 4.3.2.
Artificial Analysis, checked Oct 10, 2026
Human preference: people compare two anonymous answers and pick the better one. Ratings from the Text Arena overall board, style control on.
Arena (formerly LMArena), checked Oct 10, 2026
Agentic coding in a real terminal: can the model, driving an agent harness, finish multi step software and ops tasks. Resolution rate.
Terminal-Bench, checked Oct 10, 2026
Abstract reasoning on visual puzzles the model has not seen, semi private set. Hard to memorise, so it tracks reasoning more than knowledge.
ARC Prize, checked Oct 10, 2026
Expert level questions across many fields, no tools allowed. The Diamond set is 1,000 questions. Few models are on it yet.
Scale Labs, checked Oct 10, 2026
Median output tokens per second measured by Artificial Analysis on the provider API, for the same model setting as its index score.
Artificial Analysis, checked Oct 10, 2026
Benchmarks Anthropic reports (10)
- terminal-bench-4 (Terminal-Bench 4.0, adaptive thinking at xhigh effort (model's highest score); standard error ±2.6 pts)66.4%
- frontiercode-1-1-main (FrontierCode v1.1 (Main), adaptive thinking at max effort)54.4%
- cursorbench-4 (CursorBench 4.0, max effort)57.8%
- gdpval-aa-v2-1 (GDPval-AA v2.1 (Artificial Analysis), max effort)1846
- automationbench (Zapier AutomationBench, run and reported by Zapier during early access, no fallback models)40%
- hle-with-tools (Humanity's Last Exam, with tools, max effort)67.7%
- terminal-bench-science-0-1 (Terminal-Bench-Science 0.1, max effort (standard error ±3.5–5 pts))58.7%
- osworld-2-1 (OSWorld 2.1, partial credit)81.8%
- chartography (Chartography, with tools)89%
- chartography (Chartography, no tools)64.4%
Labs pick their own settings and harnesses, so compare these only within one lab.
Supported request parameters
What you can set when you call Claude Opus 5.5 through an OpenAI compatible API.
- Tools
tools - Structured outputs
structured_outputs - Reasoning
reasoning
Alternatives to Claude Opus 5.5
Models that take the same inputs and call tools, closest in price, cheaper ones first.
- 1.05M context
Reads text, images and files. Tool calling, structured output, reasoning. Served by 41 providers.
by openaiJul 9, 2026$4/M input tokens$20/M output tokens41 providers
- 1M context
Reads text, images and files. Tool calling, structured output, reasoning. Served by 48 providers.
by anthropicFeb 5, 2026$3/M input tokens$15/M output tokens48 providers
- 200K context
Reads text, images and files. Tool calling, structured output, reasoning. Served by 33 providers.
by anthropicJan 15, 2025$3/M input tokens$15/M output tokens33 providers
- 1.05M context
Reads text, images, files, audio and video. Tool calling, structured output, reasoning. Served by 7 providers.
by metaOct 12, 2026$4/M input tokens$20/M output tokens7 providers
- 1M context
Reads text, images and files. Tool calling, structured output, reasoning. Served by 2 providers.
by anthropicMar 29, 2026$4/M input tokens$20/M output tokens2 providers
- 400K context
Reads text, images and files. Tool calling, reasoning.
by openaiNov 12, 2025$2.50/M input tokens$20/M output tokens
More from Anthropic
- Index 43.4
Reads text, images and files. Tool calling, structured output, reasoning. Served by 26 providers.
by anthropicOct 7, 20261M context$0.10/M input tokens$0.50/M output tokens26 providers
- Index 56
Reads text, images and files. Tool calling, structured output, reasoning. Served by 26 providers.
by anthropicSep 26, 20261M context$2/M input tokens$10/M output tokens26 providers
- Index 53.4
Reads text, images and files. Tool calling, structured output, reasoning. Served by 29 providers.
by anthropicAug 29, 20261M context$10/M input tokens$50/M output tokens29 providers
- 1M context
Reads text, images and files. Tool calling, structured output, reasoning. Served by 35 providers.
by anthropicJul 23, 2026$5/M input tokens$25/M output tokens35 providers
- 1M context
Reads text, images and files. Tool calling, structured output, reasoning. Served by 38 providers.
by anthropicJun 29, 2026$2/M input tokens$10/M output tokens38 providers
- 1M context
Reads text, images and files. Tool calling, structured output, reasoning.
by anthropicJun 9, 2026$10/M input tokens$50/M output tokens
Claude Opus 5.5 questions
How much does the Claude Opus 5.5 API cost?
$4 per million input tokens and $20 per million output tokens, with cached input at $0.20. The estimator on this page turns that into a monthly bill for your own workload.
What is the context window of Claude Opus 5.5?
1,000,000 tokens of input, and it can write up to 128,000 tokens in one reply. As a rough guide, a million tokens is around 750,000 English words.
Which providers serve Claude Opus 5.5?
28 providers serve it, including Amazon Bedrock, Azure and Vertex. Prices differ by host, so the Providers table on this page lists each one with its input and output price.
Does Claude Opus 5.5 support tool calling?
Yes. It accepts tool definitions and can return JSON that matches a schema, which is what an agent needs to call your APIs.
Is Claude Opus 5.5 open source?
No. Anthropic serves it through an API only, so you cannot download the weights or host it yourself.
When was Claude Opus 5.5 released?
Sep 18, 2026, with training data up to Jun 2026.
How good is Claude Opus 5.5 compared with other models?
It scores 57.6 on the Artificial Analysis Intelligence Index and 1507 on the Arena text leaderboard. Benchmarks are a shortlist, not a verdict: test the top two or three candidates on your own tasks before you commit.
Catalog data from models.dev (MIT licensed), refreshed daily, last on Oct 15, 2026. Independent scores from my hand checked rankings, sources linked above. The summary is written from these facts.
Building on Claude Opus 5.5?
I test your real tasks on it and the two closest alternatives, then ship the agent on whichever wins on cost and accuracy.
Book Meeting