Home / Guides / Gemini CLI usage limits
Gemini CLI usage limits, plan by plan
Gemini CLI limits you by model requests per user per day: 1,500 a day on Gemini Code Assist Standard and 2,000 on Enterprise, shared by every Gemini model the CLI calls. With a paid Gemini API key, your Google Cloud project's tier sets requests per minute, tokens per minute and requests per day instead. Since June 18, 2026, Gemini CLI no longer serves Google AI Pro, Google AI Ultra or the free Code Assist for individuals: Google moved those users to Antigravity CLI.
So the 1,000, 1,500 and 2,000 a day still printed for individuals, Pro and Ultra on Gemini CLI's quota page no longer apply to anyone.
Checked 7 November 2026 against Google's Gemini CLI announcement, Gemini for Google Cloud quotas and Gemini CLI's quota page · By Soren Starck
| How you sign in | Gemini CLI still works | Limit |
|---|---|---|
| Gemini Code Assist Standard | Yes | 1,500 requests per user per day |
| Gemini Code Assist Enterprise | Yes | 2,000 requests per user per day |
| Paid Gemini API key | Yes | Your project's tier: requests per minute and per day, tokens per minute |
| Vertex AI (Gemini Enterprise Agent Platform) | Yes | Set per account, shared quota or provisioned throughput |
| Free Gemini API key | Not promised | Google's table: 250 requests a day, Flash only |
| Google AI Pro, Ultra, Code Assist for individuals | No, since June 18, 2026 | Moved to Antigravity CLI |
What counts against the Gemini CLI daily limit?
Every model request, whatever the model. Google's quota page says the daily limits are aggregated across all interactions with any model version or family, Pro and Flash alike, used with Gemini CLI or agent mode in the IDE. The count is per user, so two developers on the same Code Assist Enterprise subscription each get 2,000. Gemini CLI picks the model for each request, so a long agent task that switches between Pro and Flash still spends one pool.
There is also a per-minute ceiling. Gemini CLI's quota page says requests are limited per user per minute and depend on the service's availability when demand is high, without giving a number. Gemini for Google Cloud lists 2 requests per second per user in a project as a general limit.
Gemini CLI limits with an API key
Google's June announcement keeps Gemini CLI open to paid Gemini API keys and Gemini Enterprise Agent Platform (Vertex AI) keys. With a key, the Gemini API's rate limits apply: requests per minute, input tokens per minute and requests per day, set by your project's usage tier. They apply per project, not per key, so a second key in the same project shares the first one's limits. Gemini rate limits explained walks through what each of the three means. Vertex AI uses a dynamic shared quota or throughput you buy in advance, and its limits are specific to your account.
Gemini CLI's quota table still lists a free API key at 250 requests a day, Flash models only. The same page carries a banner saying the unpaid tier was replaced by Antigravity CLI on June 18, 2026, and Google's announcement only promises paid keys, so do not plan a project around a free key.
When do Gemini CLI limits reset?
API requests-per-day quotas reset at midnight Pacific time, which is 9:00 in Paris and 8:00 in London outside daylight saving. For Code Assist Standard and Enterprise, Google says no request goes through to any model until the daily quota resets, and gives no reset time on its quota page.
What happens when you hit the limit?
On Code Assist, Gemini CLI stops: the limit covers every model, so there is no falling back from Pro to Flash once the day is spent. On an API key, a request past a limit fails with a 429 RESOURCE_EXHAUSTED error. For anyone who runs out every day, Google's own suggestion is to move to pay-as-you-go API billing, at the per-token prices in how much Gemini costs.
Gemini CLI on Google AI Pro and Ultra: moved to Antigravity CLI
Google announced on May 19, 2026 that from June 18, 2026, Gemini CLI and the Gemini Code Assist IDE extensions would stop serving Google AI Pro, Google AI Ultra and the free Code Assist for individuals. Access through a Code Assist Standard or Enterprise licence stayed unchanged. The replacement for everyone else is Antigravity CLI, which keeps Agent Skills, Hooks, Subagents and Extensions (now Antigravity plugins), without full feature parity at launch. Gemini CLI's command reference adds that it is not optimized for Gemini 4 Argon and points to Antigravity for that model.
Gemini CLI's quota page lists Google AI Plus as not supported, and Workspace AI Standard, Plus and AI Expanded as unsupported too. The same page still lists Workspace AI Ultra at 2,000 requests a day; the June announcement does not mention Workspace plans.
What are the Antigravity CLI usage limits?
Google's announcement gives none. Antigravity bills through its own plans and credits, and the current figures, model by model, are in our Antigravity pricing guide.
How to check your Gemini CLI usage
Inside Gemini CLI, /stats model shows the current session's token counts and the limits of your quota, /stats session the session's duration and tool calls, and a usage summary prints when the session ends (command reference). Each is a snapshot of one session: none shows how much of today's quota earlier sessions used. Where else Google shows Gemini usage, API and Cloud quota pages included, is in how to check Gemini usage.
SessionWatcher keeps the day's quota in view instead. It reads the quota Google reports for your Gemini CLI sign-in and shows, for each model, the share used, what is left and when it resets, in the Mac menu bar. Gemini is part of SessionWatcher Pro. Other ways to watch it are compared in the best Gemini usage trackers.
SessionWatcherSee Gemini CLI's daily quota
before a task stops.
Native app for the Mac menu bar and the Windows taskbar. Track Claude, Codex and 20 more tools' usage, costs, and rate limits in real time.
“Fast, simple, and does exactly what it should. Definitely worth it.”
@nicojerome on GitHub
macOS 14+. Free for 7 days, no credit card.
Frequently asked questions
What are the Gemini CLI usage limits?
Gemini CLI counts model requests per user per day. Gemini Code Assist Standard allows 1,500 a day and Enterprise 2,000, across every Gemini model the CLI uses. With a paid Gemini API key the limits are your Google Cloud project's: requests per minute, tokens per minute and requests per day for its usage tier. Read on Google's pages on 7 November 2026.
Can I still use Gemini CLI with Google AI Pro or Ultra?
No. Since June 18, 2026, Gemini CLI and the Code Assist IDE extensions no longer serve Google AI Pro, Google AI Ultra or the free Gemini Code Assist for individuals. Google moved those users to Antigravity CLI, which keeps Agent Skills, Hooks, Subagents and Extensions.
When do Gemini CLI limits reset?
Gemini API requests-per-day quotas reset at midnight Pacific time. For Code Assist Standard and Enterprise, Google says requests stop until the daily quota resets but does not give the time.
What happens when I hit the Gemini CLI limit?
On Code Assist, no further requests go through to any model until the daily quota resets: the limit is shared by Pro and Flash. On an API key you get a 429 RESOURCE_EXHAUSTED error. Google suggests a pay-as-you-go API key for anyone who keeps hitting the daily limit.
How do I check my Gemini CLI usage?
Run /stats model inside Gemini CLI: it shows the current session's token counts and the limits of your quota. Gemini CLI also prints a usage summary when a session ends. Neither keeps a history across sessions.
What are the Antigravity CLI usage limits?
Google's announcement of Antigravity CLI states no limits for it. Antigravity's plans, quotas and credits are on its own pricing pages, summed up in our Antigravity pricing guide.
Sources
All read on 7 November 2026:
- Transitioning Gemini CLI to Antigravity CLI: who Gemini CLI stopped serving on June 18, 2026, and who keeps it.
- Gemini for Google Cloud quotas: Standard and Enterprise requests per user per day, what happens at the limit.
- Gemini CLI quota and pricing: the per-plan table, the June banner, /stats model.
- Gemini API rate limits: per-project limits, usage tiers, the midnight Pacific reset.
- Gemini CLI commands: /stats, /upgrade, the note on Gemini 4 Argon.