Claude Opus 5.5 API: Pricing, Model ID, and Claude Code Setup

Short answer

Claude Opus 5.5 is available on Moyi API as claude-opus-5-5 with a 1,000,000-token context window. It costs $1 per million input tokens and $5 per million output tokens (Anthropic's list price is $4 / $20). To run Claude Code on Opus 5.5 you set two environment variables — ANTHROPIC_BASE_URL and ANTHROPIC_DEFAULT_OPUS_MODEL — and paste your key. You pay per token, by card or USDT, with no subscription.

What Opus 5.5 is and when to use it

Opus is Anthropic's top model line: the strongest at reasoning and at long agentic tasks where the model plans its own steps, edits code across dozens of files, and checks the result. Opus 5.5 is the current release of that line, published on September 23, 2026.

Reach for it when a mistake costs more than the tokens: large refactors, understanding an unfamiliar codebase, or any task where the whole repository has to stay in context (a million tokens is roughly 30–40k lines of code plus docs). For routine edits and completions it is cheaper to keep Sonnet 5 or Fable 5.1 as the default and switch to Opus 5.5 with /model only for the hard task — setup below.

Pricing: list vs. Moyi API

Per 1 million tokens, in USD:

The live number is always on the pricing page; it moves with the upstream price list, so the figures here are as of the publish date.

What that means in practice: a typical Claude Code session on a mid-size project reads about 200k input tokens and writes about 20k. At list price that is $0.80 + $0.40 = $1.20; on Moyi API it is $0.20 + $0.10 = $0.30. When the project doesn't change between requests, most of the input is billed at the cache-read rate and the session gets cheaper still — see Billing FAQ for how caching is charged.

Set up Opus 5.5 in Claude Code

You need a Moyi API API key — create one in the console. If Claude Code isn't installed yet, start with the installation guide, which covers macOS, Windows, and Linux.

macOS and Linux

Add to ~/.zshrc (or ~/.bashrc if you use bash):

export ANTHROPIC_BASE_URL=https://api.moyiapi.com
export ANTHROPIC_AUTH_TOKEN="your Moyi API key"
export ANTHROPIC_MODEL=claude-opus-5-5
export ANTHROPIC_DEFAULT_OPUS_MODEL=claude-opus-5-5
export ANTHROPIC_DEFAULT_SONNET_MODEL=claude-sonnet-5
export ANTHROPIC_DEFAULT_HAIKU_MODEL=claude-haiku-4-5

Then source ~/.zshrc and run claude.

Windows (PowerShell)

setx ANTHROPIC_BASE_URL "https://api.moyiapi.com"
setx ANTHROPIC_AUTH_TOKEN "your Moyi API key"
setx ANTHROPIC_MODEL "claude-opus-5-5"
setx ANTHROPIC_DEFAULT_OPUS_MODEL "claude-opus-5-5"
setx ANTHROPIC_DEFAULT_SONNET_MODEL "claude-sonnet-5"
setx ANTHROPIC_DEFAULT_HAIKU_MODEL "claude-haiku-4-5"

setx writes the variables permanently; close and reopen the terminal for them to take effect.

Verify and switch inside a session

In a running Claude Code session type /model. The list should show opus, and it resolves to claude-opus-5-5 because we pinned it with ANTHROPIC_DEFAULT_OPUS_MODEL. The same command lets you drop to sonnet for cheap work and come back to opus when the task deserves it. The active model is shown in the status line at the bottom of the window.

To pin the setting for one project instead of the whole machine, put it in .claude/settings.json at the repository root:

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.moyiapi.com",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "claude-opus-5-5"
  }
}

Calling the API directly

The same key works outside Claude Code — from scripts, a backend, any client. Two formats to choose from.

Anthropic format (what the official SDK uses):

curl https://api.moyiapi.com/v1/messages \
  -H "x-api-key: YOUR_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-opus-5-5",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Explain what this code does: ..."}]
  }'

OpenAI-compatible format — if you already have code written against the OpenAI SDK, only the base URL, key, and model name change:

from openai import OpenAI

client = OpenAI(base_url="https://api.moyiapi.com/v1", api_key="YOUR_KEY")
resp = client.chat.completions.create(
    model="claude-opus-5-5",
    messages=[{"role": "user", "content": "Write a test for parse_date"}],
)
print(resp.choices[0].message.content)

Full parameter reference: API integration.

Paying for usage

You top up a balance in the console and are charged only for tokens actually used — there is no subscription or monthly fee.

• Card — Visa, Mastercard and other major cards via Stripe.
• USDT — choose "Pay with crypto" in the top-up dialog, send USDT to the address shown; the balance is credited within one to five minutes.

There is no minimum bundle and nothing expires: top up $10 and you spend exactly $10 on tokens at the prices in the table above.

FAQ

/model has no opus entry, or it points at an older model. Check that ANTHROPIC_DEFAULT_OPUS_MODEL is set in the shell that launched Claude Code (echo $ANTHROPIC_DEFAULT_OPUS_MODEL). On Windows, open a new terminal after setx.

How is claude-opus-5-5 different from claude-opus-5? They are two different models with different IDs; Claude Code will not move you to 5.5 on its own. If your settings pin claude-opus-5, replace it.

How do I enable the 1M context? Nothing to enable — the context limit is a property of the model. You do pay for every input token, so on large repositories watching the cache matters more than watching the limit.

What are cache read and cache write in the price table? When the same long context (your repository, say) is sent again, it is served from cache at the read rate — several times cheaper than normal input. The first send is billed as a write. Details in the billing FAQ.

Are there request limits? Yes, per model — see the rate limits page.

Next steps

• Install and configure Claude Code — if you're starting from scratch.
• What Claude Code really costs — a breakdown on typical workloads.
• Create an API key — and run Opus 5.5 in five minutes.