A bunch of you have been asking for transparency around Codex CLI (and now VS Code extension) rate limits. We just published our first ranges: - Average Plus/Team users can send 30-150 messages every 5 hours - Average Pro users can send 300-1,500 messages every 5 hours Those
@embirico There’s a small typo in the Codex CLI page: “After installing, you’ll can find” https://developers.openai.com/...
@lai_thanh_trung thanks pushed a fix
@embirico To be clear by message you mean a prompt I type into codex, not a file interaction?
@embirico ive hited limits manually as a pro user 😂😂😂
@embirico Enterprise rate limit = 0 to keep things simple
@embirico Usage scales with expertise. Interesting how productivity translates into message volume. This incentivizes efficient prompting.
@embirico any plans on a middle level plan around $100 to have limits some where between them?
@embirico Already reached the limit usage in 1.5 days. It's a good product but the usage limit is too low, and Pro sub is just too much. Why not add another tier that is more affordable than PRO but has higher usage limits than PLUS ?
@embirico How's the reasoning level play into this
@embirico Can you clarify whether 30-150 messages is unique inputs from the user, or is it possibly eaten up by multiple requests being made under the hood? Why not just give a reasonable weekly limit with much higher 5hr limits? The worst is getting deep into a prob and cut off.
@embirico You need a 100usd plan, you can call it max Plus10x. 10x, compared to normal plan!
@embirico So pack more instructions per message 🤫😉 GPT5 is great at long-horizon instruction following compared to other SOTA models. So.. Don’t: - Read codebase - Read docs - Review issues - Plan feature implementation - Go code Do: - Review codebase and docs in /docs. Then
@embirico it’s kinda low
@embirico Teams with x number of licenses should have more limits then plus, as teams purchased for multiple employees to enhance their work
@embirico Bro, literally just give me a bar across the top that fills up as I go, shows me where I'm at, and when it will next reset.
@embirico I'm on the Plus plan. About ten days ago I tried the Codex CLI and hit a usage cap just 12 minutes in—speechless. It's less painful now. In Japan, more folks are seeing how well Codex performs. Keep up the great work!
@embirico Hmm so currently on my Plus plan I get about 10-30 minutes of "actual" use out of it with reasoning_effort=low, actual here meaning not one-off things like "update this file", but agentic feature implementations The Pro plan would 10x that, so 100-300min (5h top end) At least
@embirico What about an intermediate tier?
@embirico Plus plan with Codex VSCode: Got the 5h limit after a day, no big deal. But one day later I hit another limit (this is only the second time I've hit any kind of limit), now it asks me to wait for 5 days! 5 days, no kidding! The weekly limits are unworkable. Pls fix the limits!
@embirico I have the pro subscription and got this in the extension, I'm a normal user, how's this possible?
@embirico Love the new models, but this kinda sucks
@embirico Why don’t you sell “sessions”? Like 10$ to unlock/reset weekly limits for example
@embirico can you bring /slash commands to the vs code extension too pls
@embirico I am rate limited continuously. I am not scripting at all, but I find it fairly easy to keep up with 5-7 parallel instances following implementation plans I built with acceptance criteria. Mostly not vibe coding as much as supervising implantation. Would pay much more. Let me!
@embirico @doodlestein in case you didn't see. this is pretty sweet..
@embirico we need a 100$ plan like claude for limits somewhere in the middle pro is overkill for me and plus will be useless
@embirico Please bring 50$, and 100$ plan
@embirico Why can't you just tell us what is the equivalent token spending and allow us to track it both weekly and hourly?
@embirico love the transparency. would love to see the weekly limits too
@embirico Please do $40 plan
@embirico stream error: unexpected status 404 Not Found: ; retrying 1/5 in 212ms… ⚠ stream error: unexpected status 404 Not Found: ; retrying 2/5 in 434ms… ⚠ stream error: unexpected status 404 Not Found: ; retrying 3/5 in 774ms… i keep getting this error
@embirico Could you have an easy way to switch to API usage when we hit limits on a plan?
@embirico What are weekly limits