Opus 5.5 is native, nothing to translate
Opus 5.5 is Anthropic's own flagship, served over the Anthropic Messages API that Claude Code was built for. So this is the easy case: no proxy, no OpenAI-compatible shim, usually no custom base URL. You authenticate with Anthropic and pick Opus 5.5.
The config
What to set
- AuthSign in with your Claude plan, or set ANTHROPIC_AUTH_TOKEN (ANTHROPIC_API_KEY) to a valid Anthropic key. No custom ANTHROPIC_BASE_URL needed.
- ModelSelect the Opus 5.5 model using the exact model string Anthropic's API exposes.
- Mind the costOpus is the premium tier. Watch token usage on long agent loops; consider a cheaper model for routine edits.
When Opus 5.5 is worth it
Hard problems
Opus is the heavy tier, best for gnarly debugging, large refactors and multi-step reasoning where a cheaper model stalls.
Native fit
It's Anthropic's own flagship, so Claude Code's agent loop and tool-calling are tuned for it.
Cost tradeoff
It's the most expensive tier; for routine edits and boilerplate, a Sonnet or cheaper model is often enough.
The cleaner option: Opus for the hard stuff, cheaper for the rest
Opus is native and easy to select, but Claude Code is single-provider by design, so mixing Opus for hard tasks and a cheaper model for routine work means editing env vars and restarting each time. An agent where the model is a setting lets you keep Opus 5.5 on tap and drop to Sonnet, DeepSeek or a local model in a click.
What that looks like with Wuwei
Opus 5.5 in the model list
Pick it in the UI; Sonnet, GPT, DeepSeek, Kimi, Qwen and OpenAI-compatible endpoints sit alongside.
One-click switch
Use Opus for the hard problem, then drop to a cheaper model for the cleanup, no config edit, no restart.
Free, open source, local
MIT-licensed, runs on your machine, reads your real files; you only pay your own provider usage.
FAQ
Does Claude Opus 5.5 work with Claude Code natively?
Yes. Opus 5.5 is an Anthropic model on the Anthropic Messages API that Claude Code is built for, so it works natively, no translating proxy and no custom endpoint. Authenticate with your Claude plan or Anthropic key and select Opus 5.5.
Opus 5.5 vs Sonnet 5.5, which should I use in Claude Code?
Opus is the heavier, pricier tier for hard reasoning and big refactors; Sonnet is the balanced value pick for everyday work. Many people keep Opus for tough problems and use Sonnet for routine edits.
Do I need a proxy for Claude Opus 5.5 in Claude Code?
No. Proxies are only for non-Anthropic models. Opus 5.5 is Anthropic's own flagship, so Claude Code talks to it directly.
How do I switch between Opus 5.5 and a cheaper model?
Use an agent where the model is a setting. Wuwei is free and open source, keep Opus 5.5 for hard tasks and drop to Sonnet, DeepSeek or a local model in one click instead of editing env vars.