Running Cursor on Unlimited Claude Access

Cursor lets you bring your own API key, which means you can run it on a flat-rate unlimited Claude plan and stop thinking about per-request cost. Here is how to wire it up and why heavy Cursor users come here for it.

Bring your own key into Cursor

Open Cursor Settings, go to Models, and in the OpenAI API Key section enable Override OpenAI Base URL. Paste the gateway base URL ending in /v1 (Cursor appends /chat/completions) and your unlimited key in the field below — the gateway forwards the request to Claude.

Add a Claude model with + Add Model, click Verify, then open a chat and select it. Cursor's custom-key path uses the OpenAI-compatible format, which is exactly what the gateway speaks.

Why heavy Cursor use suits unlimited

Cursor's chat, edits, and back-and-forth refinement all hit the model, and a productive day can be hundreds of requests. On per-token billing that is a meter you watch; on a flat-rate unlimited plan it is simply included.

If you live in Cursor — long sessions, big files, lots of iteration — unlimited access removes the friction of rationing usage. You set the override once and use Claude as hard as your work demands for one predictable price.

What unlimited does and does not cover

Within your plan there are no token limits and no per-token charges — only fair-use rate limits that keep the gateway responsive. Cursor's own Tab autocomplete always uses Cursor's built-in model, not your gateway, so that part is unaffected either way.

For chat, edit, and the agentic features that do route to your key, unlimited access means you can keep iterating without counting tokens — which is the whole reason to run Cursor this way.

# Cursor → Settings → Models → OpenAI API Key
# Override OpenAI Base URL: https://claudeapikey.dev/v1
# OpenAI API Key: <your unlimited key>
# + Add Model: claude-sonnet-4-5  → Verify → Save
# Open a chat, pick the model, code without watching a meter.

Frequently asked questions

Can I run Cursor on an unlimited Claude plan?
Yes. Cursor supports bring-your-own-key — enable Override OpenAI Base URL, point it at the gateway's /v1 root, paste your unlimited key, add a Claude model, and verify.

Why is unlimited good for Cursor?
Heavy Cursor use is hundreds of requests a day across chat and edits. Flat-rate unlimited includes all of that for one price instead of a per-token meter.

Does unlimited cover Cursor Tab autocomplete?
Cursor Tab always uses Cursor's own built-in model, not your gateway — so it is unaffected. Unlimited covers the chat/edit features that route to your key.

Are there any limits on the unlimited plan?
No token limits and no per-token charges during the plan — only fair-use rate limits to keep the gateway fast.

Claude API Key is an independent API gateway. It is not affiliated with, endorsed by, or a reseller of Anthropic. Claude and related model names are trademarks of their respective owners.