Claude Opus 5.5 costs $4 per million input tokens and $20 per million output tokens on Roar AI. Anthropic’s recommended default for demanding work, and cheaper than the Opus it replaced.
Example: 1,000 requests with 3,000 input and 1,000 output tokens each cost $32. Every rate is published and stays the same from one request to the next. See every model.
Rates as of 20 November 2026.
claude-opus-5-5For long, open-ended work that needs judgement throughout, yes. Anthropic says that is where Opus is clearly stronger than Sonnet 5.5, and it scored higher than Fable 5.1 on professional knowledge work in Anthropic’s own GDPval-AA test, 1846 against 1735.
For most other work, try Claude Sonnet 5.5 first. It nearly matches Opus on most measured work at a lower rate, and it beat Opus on Terminal-Bench 4.0. Anthropic itself cautions that benchmark leads have become a weak guide to real differences, so compare the two on your own tasks.
| Model | Input | Output | Choose it for |
|---|---|---|---|
| Claude Opus 5.5 | $4 | $20 | Complex coding and knowledge work. |
| Claude Sonnet 5.5 | $2 | $10 | Most everyday coding and documents, at a lower rate. |
| GPT-6.1 Sol | $2 | $10 | Production agents, when cost per task decides it. |
| Claude Fable 5.1 | $10 | $50 | Problems Opus cannot solve. |
Prices per 1M tokens.
Opus 5.5 replaced Opus 5 in September 2026 and is the model Anthropic tells most people to start with. Anthropic puts it at about 40% cheaper than its predecessor on typical work and more than 30% faster at writing its answer. On professional knowledge work it scored higher than Fable 5.1 in Anthropic’s own GDPval-AA test (1846 against 1735), and it reached 66.4% on Terminal-Bench 4.0.
Thinking is always on, at medium effort by default. For long coding sessions, reviewing a large codebase and business analysis it is a dependable choice for demanding work; Fable 5.1 is worth the extra cost only on problems Opus cannot solve.
Use the OpenAI SDK you already have: point it at Roar AI and set the model name. Read the docs.
from openai import OpenAI
client = OpenAI(base_url="https://api.roar-ai.com/v1", api_key="roar_live_…")
reply = client.chat.completions.create(
model="claude-opus-5-5",
messages=[{"role": "user", "content": "Summarise this contract in plain English."}],
)Input is $4 per 1M tokens; output is $20 per 1M tokens. For example, 1,000 requests with 3,000 input and 1,000 output tokens each cost $32.
Yes. Every model on Roar AI is on the same API key and the same monthly invoice, so switching from Claude Opus 5.5 to another model only means changing the model name in your code.
No. Anthropic does not publish the weights, so it is only available through an API like this one.
Claude Opus 5.5 can read up to 1M tokens in one request, and write an answer of up to 128K tokens.
Yes. Claude Opus 5.5 reads images and PDFs directly.
Create a free account, get an API key and call Claude Opus 5.5 with the OpenAI SDK.