Antigravity Pro Users Hit Multi-Day Quota Lockouts
Google Antigravity Pro subscribers report 5-7 day lockouts on premium models after quota changes replaced five-hour refreshes with weekly caps and AI credit overages, sparking backlash across developer forums.
5 min read
Google Antigravity Pro subscribers are reporting multi-day lockouts on premium models after the company tightened quota enforcement and introduced AI credit overages. Developers who were previously hitting five-hour refresh cycles on models like Gemini 3.1 Pro and Claude Sonnet 4.6 are now being locked out for five to seven days once they hit a weekly cap that wasn't prominently advertised. The changes push heavy users toward the $250/month Ultra plan or paid AI credit top-ups.
TL;DR
- Antigravity Pro ($20/month) now enforces weekly caps that lock premium models for 5-7 days once exhausted
- Previous system allowed five-hour refreshes with more generous weekly limits
- A cross-model bug locks all models simultaneously, even ones the user never touched
- Ultra ($250/month) has no weekly cap - the only fix is a 12.5x price jump with no middle tier
- Developers are switching to Cursor, Claude Code, and GitHub Copilot
What Changed
When Antigravity launched in November 2025, Pro subscribers got five-hour quota refreshes with weekly limits generous enough that most developers never hit them. The system has since tightened through a series of changes between January and March 2026.
The current state: Pro users still see five-hour refreshes, but only until they hit a weekly cap. Once the weekly limit is reached, models lock for the remainder of the week - typically five to seven days. Some developers report hitting the weekly cap after as little as 20 minutes of intensive usage, then facing lockouts of up to 167 hours.
One developer on Google's AI forum documented the drop: "Before January I could use over 300 million input tokens and 1-2 million output in a week for the Gemini Pro models. This week I hit my weekly rate limits at less than 9 million input and 200 thousand output tokens."
A separate bug compounds the problem. Exhausting the quota on one model - say Claude Sonnet 4.6 - can simultaneously lock all other models, including ones the user never called. Forum users report being blocked from GPT-OSS 120B despite never using it.
The Pricing Gap
Google's answer to the lockouts is the Ultra plan at $249.99/month, which removes weekly caps entirely. The 12.5x price jump from Pro ($20) to Ultra ($250) with no intermediate tier has been called "predatory" by developers on the forums.
| Tier | Cost | Quota |
|---|---|---|
| Free | $0 | Weekly-based limits, hits after 2-3 hours |
| Pro | $19.99/month | 5-hour refresh until weekly cap, then 5-7 day lockout |
| Ultra | $249.99/month | 5-hour refresh, no weekly cap |
The AI credits system, introduced as a way to continue past quota, charges Vertex API rates. Developers can buy 2,500 credits for $24.99 or 20,000 for $199. The exact credit-to-usage conversion for Antigravity sessions is not publicly documented, making it difficult to budget.
Developer Reactions
The Google AI Developers Forum has become the primary venue for complaints:
- "With the current quotas and lockouts it honestly feels like I'm using a free plan" - Pro subscriber
- "I paid for the service but I can't even use it to develop a full small project" - Petra_Christalbert
- "Ran out of quota in less than 30 minutes. I'm cancelling the plan because that's just ridiculous" - forum user
- "People who have been on AI Pro plan since Antigravity was released feel that this is basically a bait and switch" - forum user
Developers are comparing the situation unfavorably to competitors. Cursor at $20/month offers unlimited usage on its Pro tier. Claude Code Max runs $100-200/month with reliable Opus 4.6 access. GitHub Copilot costs $19-39/month. All offer more predictable usage than Antigravity Pro's current state.
Google's Response
Google has acknowledged bugs but framed the quota changes as improvements. The official Antigravity account posted on X: "We're evolving Google AI plans to give you more control over how you build. Every subscription includes built-in AI credits, which can now be used for Antigravity, giving you a seamless path to scale."
A Google blog post describes Pro and Ultra subscribers as having "priority access, featuring our highest, most generous rate limits with quotas that refresh every five hours" - language that omits the weekly cap that overrides those refreshes.
On the forum, a Google representative confirmed that "weekly limits apply for all models" on Pro and that "these limits do not apply to Google AI Ultra, which continues to be the best plan for power developers." A separate representative apologized for bugs and confirmed escalation to internal teams, but no timeline for fixes has been given.
The Register asked Google directly to clarify what changed and what an AI credit buys when used with Antigravity. Google did not respond.
The Bigger Picture
Antigravity launched as Google's answer to the agentic IDE wave - a free, powerful alternative to Cursor and Windsurf backed by Gemini 3 and third-party models. The generous initial quotas drove rapid adoption. Tightening those quotas four months later, with the only relief being a 12.5x price increase or opaque credit purchases, erodes the trust that drove that adoption.
The pattern isn't unique to Google. Every AI-powered developer tool faces the tension between subsidized growth and sustainable unit economics. But the gap between Antigravity's marketing ("most generous rate limits") and the lived experience of Pro subscribers (20 minutes of usage triggering a week-long lockout) is unusually wide. Developers who built workflows around Antigravity's initial generosity are now evaluating whether to pay $250/month, buy credits, or move to a competitor that charges less and delivers more predictably.
Sources:
