I don't know if they should be handing out more credits right now considering that my Sonnet requests in Claude Code are routinely delayed by several minutes, presumably due to capacity issues...
> I suspect that there's an unstated additional qualification: that your account hasn't previously received an extra usage credit.
My account has never received any credit. I subscribed to Pro only a few weeks ago.
I received extra usage credit in February, and I got the email today asking me to claim this extra credit. Managed to claim this one too.
I wouldn't be surprised if they pulled it back so they could spread out the load.
edit: whoops, meant to leave this as a reply to the (now-sibling) comment from 'flutas.
If you haven't, then you don't get the credit.
Toggling it again to turn it off immediately after (which I assume most people will miss) seems to guard against this.
I'd actually be happy with this if they turned OpenCode support back on.
As in, what you're toggling on is "use my extra usage balance" not "automatically spend money by adding to my extra usage balance".
[x] Turn on extra usage to keep using Claude if you hit a limit.
^^^ This toggles whether it will automatically spend from your pre-paid balance when you hit limits. $20
Monthly spend limit
^^^ I'm not actually sure which this toggles, but I think it is either a maximum amount to reload per month, OR I think it is a maximum amount of your balance to spend per month. I think it's how much to auto-reload but it is a confusing setting. Auto-reload off
^^^ This toggles whether it auto-reloads. It accepts a trigger balance and how much to top it off to. This setting shows the following text, which is You agree that Anthropic will charge the card you have on file in the amount above on a recurring basis whenever your balance reaches the amount indicated. To cancel, turn off auto-reload.Fin isn't half as useful as Claude, and if that's not telling, I don't know what would be.
- https://www.reddit.com/r/ClaudeAI/comments/1s7mkn3/psa_claud...
- https://www.reddit.com/r/ClaudeAI/comments/1s7zgj0/comment/o...
But I think it only lasted for a short period of time.
As far as I can tell, they now have Free, Pro, Max x5, Max x20, Team, and Enterprise tiers. Plus session usage, weekly usage, extra usage (up to a spend limit (set by you) and/or a monthly cap (set by Anthropic)), and now usage bundles, which are extra usage but with a lower price and a fixed amount.
I don't know, but it really feels like there's an LLM churning away somewhere, endlessly tinkering on a PRICING.md file that is gradually accreting strategic slop.
Wow, that's shitty Antropic.
I am guessing the vast majority of eligible credit will not be claimed or expire unspent.
Worth mentioning that normal credit expires after 12 months too, use it or lose it.
Anthropic recently (within the last few months) gave out what amounted to around $70 CAD im free credit. I rationed it slowly, spending maybe $5 in a month, using it a few dozen times to allow a Haiku-based task to finish past my usage window.
When Opus and Sonnet 4.6 released, I made the unfortunate mistake of "experimenting" with them on some work that couldn't be thrown away. I hit my timed usage cap, allowed Claude Code to consume 'extra usage' credits, and... nearly vaporized the entire credit balance within a couple tasks.
I understand that Opus and Sonnet are (considerably) more expensive than Haiku, but watching money burn by the dollar, in real-time, was enough for me to turn off extra usage entirely.
I also really dislike compacting though, and perhaps that is the catalyst for why i never burn unexpected amounts. In my Claude.md i have instructions to wrap up work around 150k tokens, from there i usually have enough of a window of tokens to decide what to do with the work done that session.
sometimes thats a simple ya, in those situations i use the remaining usage to have claude update .claude/ with relevant changes, if i've got more wiggle room for tokens, i might even craft up my next plan which is usually TDD style, and i have it write the failing tests before ending the session and passing the handoff.md which is the finished plan to a fresh context window.
This is probably not the best way for rapidly getting stuff done, but i think its one of the best ways to be effective with less usage at your disposal.
EDIT: I guess that means they are indeed running out of cash.