[BUG] API Error: Rate Limit when using claude code (vscode) on max plan

Found 3 possible duplicate issues:

1. https://github.com/anthropics/claude-code/issues/29579
2. https://github.com/anthropics/claude-code/issues/30723
3. https://github.com/anthropics/claude-code/issues/32536

This issue will be automatically closed as a duplicate in 3 days.

• If your issue is a duplicate, please close it and 👍 the existing issue instead
• To prevent auto-closure, add a comment or 👎 this comment

🤖 Generated with Claude Code

Sorry to hear about this frustrating experience — being rate-limited despite being on the max plan is definitely unacceptable.

A few things that might help in the meantime:

1. Check VS Code extension vs Claude Code CLI: The rate limits may be separate. Try using the CLI directly (claude) instead of the VS Code extension to see if the issue persists.

1. API key vs Anthropic account: If you're using an API key directly (vs. authenticated via anthropic.ai), check if there's a separate rate limit on that key at the Anthropic console.

1. Token usage monitoring: Sometimes high token usage triggers limits. Check your console for detailed usage.

For a more robust solution, you might want to look into agent frameworks with local processing (like OpenClaw) that can run autonomously without hitting API rate limits for routine operations — they only call the LLM when needed, reducing overall API consumption.

Would be happy to help debug further if you share more details about your setup!

I have the same problem with the VScode extension since yesterday. I tried 2 different machines, tried logging in with the Claude MAX subscription as well as the API key, ran in empty folder, restarted and reinstalled multiple times and still get the same error. Everything works in terminal and desktop.

The error message is misleading. In my case, “API Error: Rate limit reached” was triggered because the VS Code chat had switched to a 1M context model that is not included in my Pro plan. This was not a real rate limit issue. The message should clearly indicate that the selected model is unavailable for the current plan, instead of suggesting usage exhaustion.

I got same problem today, didn't have this issue a few days ago. The Windows Claude app on code tab works fine though, so its a bug specific to VSCode,

@gabriel-vthereagency thats genious, I simple went to the VSCode extension model select menu, hit just "Sonnet 4.6" and now it works!

So the bug seems to be related to having an invalid model selected somehow, and then the rate limit error is misleading.

It should say "Invalid model for your plan" or something to prompt people to try switching models to fix.

Same issue here. Pro plan, Windows 11, Cursor (VS Code fork), latest Claude Code extension.

Opus 4.6 instantly returns "API Error: Rate limit reached" on any message, even a simple "hello". Sonnet works fine in the same extension. Opus also works perfectly in Claude Desktop (Code tab) and on claude.ai — so the issue is specifically with the VS Code extension + Opus.

My usage is at 5% session / 14% weekly, nowhere near any limit.

I've tried everything:

• Full uninstall of the extension
• Deleted %USERPROFILE%\.claude folder
• Fresh reinstall + re-login with Claude.ai Subscription
• No ANTHROPIC_API_KEY set
• Set CLAUDE_CODE_DISABLE_1M_CONTEXT=true (no effect on the extension)
• Contacted support — they said all incidents were resolved and it was on my end, but nothing fixed it

Also noticed the Account & Usage panel shows "Plan: Claude api" instead of "Pro", which seems wrong.

It's been a week since this issue was opened. @csidi90 is it still happening for you?

See also #30824 and #33154 — same pattern. This needs attention, paying subscribers can't use Opus through the extension.

Really looking forward to hearing news from this issue.

I am at same point than most of people here....

Fix: In VS Code/Cursor, go to Extensions → Claude Code → click the dropdown arrow next to Uninstall → "Install Another Version" → select v2.1.72. Disable auto-update to prevent it from updating back.

Fix: In VS Code/Cursor, go to Extensions → Claude Code → click the dropdown arrow next to Uninstall → "Install Another Version" → select v2.1.72. Disable auto-update to prevent it from updating back.

Thanks a lot buddy! it’s working perfectly with that temporary solution, think that Claude should fix that soon…

We're coming up with a fix! Please use /model to select the opus model if you see this again

You need to put CLAUDE_CODE_DISABLE_1M_CONTEXT=1 in env variables for opus.

You can switch to sonnet via CLI /model and ask it to change that for OPUS settings and you're good to go.

same issue here, I just changed the model to Haiku then to Sonnet and it workssss :)

Switching to v2.1.72 did not help. Any other thought?
Also, tried selecting specific model, still keep getting the error.

On checking usage ? - a lot reamining.

I am seeing the same thing in VS Code. Switch models away from [1m[ appears to have solved it but that's hardly the point. I'm on Max x5 at the moment, I've reauthenticated and the context shows 1m but get API rate error trying anything.

I was having the issue, but I updated to v2.1.81, and it seems to have fixed it (at least for now).

I downgraded from 2.1.81 to 2.1.79 and it helped. Never used Opus, API rate limit happened with Sonnet

I was having the issue, but I updated to v2.1.81, and it seems to have fixed it (at least for now).

Update: the issue is back. Toggling between opus 200k and opus 1m fixes the issue.

the issue is back. Toggling between opus 200k and opus 1m fixes the issue. --- How to do this?

Also, in VS Code extension do we need some more access/billing? I have the pro version.

Downgraded to vscode extension 2.1.81 fixed it.
Still selected Opus as model

This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.