Head to head · live data
ChatGPT + Codex vs Google AI (Gemini, Antigravity): plan limits and pricing
Every plan of both tools side by side: price, how usage is metered, when it resets and each limit the vendor states, copied from the vendor's own page. Where a vendor publishes no number, the row says so.
14 plans. Checked October 30, 2026. Full pages: ChatGPT + Codex · Google AI (Gemini, Antigravity) · change log
OpenAI
ChatGPT + Codex
- Free plan
- Free
- Entry paid plan
- Go, $8/mo
- Top self-serve tier
- Pro ($500), $500/mo
- Metered as
- messages per 5 hours
- Reset windows
- five-hour window, week
- Numbers published
- 30 of 40 limits
Per the vendor, the number of Codex messages depends on the model, task size and complexity, and whether work runs locally or in the cloud. Plus and Standard Business get local-message estimates per five-hour period, given as a range per model; these are "not fixed message limits", and the usage dashboard shows current limits and reset times. Cloud tasks may use more of the allowance than local messages. Local messages and cloud chats share the plan's usage allowance, and weekly limits may also apply. Pro plans currently have no five-hour limit. ChatGPT Work shares the same pricing, credits and usage limits as Codex. Fast mode uses included usage at 2.5x (2x on purchased credits). Ultrafast uses 8x (6x on credits). Image generation uses limits about 3-5x faster. After reaching limits, Plus/Pro users and flexible-pricing workspaces can buy credits, which are billed per 1M tokens per model. Enterprise/Edu with flexible pricing have no fixed rate limits.
- Free plan
- Free
- Entry paid plan
- AI Plus, $4.99/mo
- Top self-serve tier
- AI Ultra 20x, $199.99/mo
- Metered as
- usage allowance
- Reset windows
- five-hour window, week, month
- Numbers published
- 19 of 27 limits
In the Gemini app, Google meters compute used instead of daily prompts. Usage limits 'factor in the complexity of your prompt, the features you use, and the length of your chat. Your limit refreshes every 5 hours until you reach your weekly limit.' If you hit the cap on the biggest models, you are moved to smaller models. Media generation, Deep Research, the Pro model and Extended thinking/Deep Think use up the limit faster, and so do higher effort levels. Tiers are multiples: Plus is 2x and Pro 4x the 'standard' (no-plan) limit, and Ultra is 5x or 20x Pro. No prompt counts are published. Google Antigravity (app, IDE extensions and CLI) has its own agent quota that 'correlate[s] with the amount of work done by the agent'. Pro and Ultra get a quota refreshed every five hours up to a weekly rate limit; users without Pro or Ultra get a quota refreshed weekly. No figures are published. Pro and Ultra subscribers can buy AI credits for overage in Antigravity and Flow, and the subscriptions page now says credits can also extend Gemini app limits. Since June 18, 2026, Gemini CLI has been replaced by Antigravity CLI for unpaid and Google One users, so the old Gemini CLI daily request counts no longer apply to these plans.
What differs
Stated from the two plan sets, no ranking: which tool fits depends on which window and unit matches your work.
- Entry price: ChatGPT + Codex Go at $8/mo; Google AI (Gemini, Antigravity) AI Plus at $4.99/mo.
- Top self-serve tier: ChatGPT + Codex tops out at Pro ($500), $500/mo; Google AI (Gemini, Antigravity) tops out at AI Ultra 20x, $199.99/mo.
- ChatGPT + Codex has a free plan (Free); Google AI (Gemini, Antigravity) has a free plan (Free).
- Metering unit: ChatGPT + Codex counts "messages per 5 hours", Google AI (Gemini, Antigravity) counts "usage allowance".
- Reset windows: ChatGPT + Codex five-hour window, week; Google AI (Gemini, Antigravity) five-hour window, week, month.
- ChatGPT + Codex prices 1 plan per seat; Google AI (Gemini, Antigravity) lists no per-seat plan.
- Running out: ChatGPT + Codex: extra usage, Can purchase additional credits after reaching usage limit; Google AI (Gemini, Antigravity): extra usage, Purchasable AI credits for Antigravity overage, Flow and the Gemini app.
- Transparency: ChatGPT + Codex publishes a number for 30 of 40 limits; Google AI (Gemini, Antigravity) publishes a number for 19 of 27 limits.
Plan by plan
| Tool | Plan | Price | Metered as | Headline limit | Window | Source |
|---|---|---|---|---|---|---|
| ChatGPT + Codex | Free | Free | messages per 5 hours | Codex: GPT-6 Luna at Standard speed in the desktop app, subject to rollout; no message figures publishedno number published | five-hour window | Codex pricing |
| Go | $8/mo | messages per 5 hours | Codex: GPT-6 Luna at Standard speed in the desktop app, subject to rollout; no message figures publishedno number published | five-hour window | Codex pricing | |
| Plus | $20/mo | messages per 5 hours | Codex local messages, GPT-6 Astra: 5 to 45 per 5 hours | five-hour window | Codex pricing | |
| Pro ($100) | $100/mo | usage allowance | Codex 5-hour limit: None (vendor: Pro plans currently have no five-hour limit) | five-hour window | Codex pricing | |
| Pro ($200) | $200/mo | usage allowance | Codex 5-hour limit: None (vendor: Pro plans currently have no five-hour limit) | five-hour window | Codex pricing | |
| Pro ($500) | $500/mo | usage allowance | Codex 5-hour limit: None (vendor: Pro plans currently have no five-hour limit) | five-hour window | Codex pricing | |
| Business | Per user per month when billed monthly; 2+ users$20/mo annual | messages per 5 hours | Codex local messages, GPT-6 Astra: 5 to 45 per 5 hours | five-hour window | Codex pricing | |
| Business Codex (pay as you go) | No fixed seat fee; usage-based | usage (pay as you go) | Usage: Pay as you go based on usage; no limit figures publishedno number published | Chatgpt Pricing | ||
| Enterprise & Edu | Custom pricing; contact sales | credits | Rate limits (flexible pricing): No fixed rate limits; usage scales with credits | Codex pricing | ||
| Google AI (Gemini, Antigravity) | Free | Free | usage allowance | Storage: 15 GB | Google AI subscriptions | |
| AI Plus | $4.99/mo | usage allowance | Gemini app usage: 2x the Free limits | five-hour window | Google AI subscriptions | |
| AI Pro | $19.99/mo | usage allowance | Gemini app usage: 4x the Free limits | five-hour window | Google AI subscriptions | |
| AI Ultra 5x | $99.99/mo | usage allowance | Gemini app usage: 5x the AI Pro limits | five-hour window | Google AI subscriptions | |
| AI Ultra 20x | $199.99/mo | usage allowance | Gemini app usage: 20x the AI Pro limits | five-hour window | Google AI subscriptions |
One headline limit per plan. Every limit, with its wording and source, is on the ChatGPT + Codex and Google AI (Gemini, Antigravity) pages.
ChatGPT + Codex vs Google AI (Gemini, Antigravity): common questions
Which is cheaper, ChatGPT + Codex or Google AI (Gemini, Antigravity)?
Google AI (Gemini, Antigravity) has the lower entry price. ChatGPT + Codex Go is $8/mo; Google AI (Gemini, Antigravity) AI Plus is $4.99/mo. Prices from the vendors' pricing pages, checked 2026-10-30.
How do ChatGPT + Codex and Google AI (Gemini, Antigravity) count usage?
ChatGPT + Codex: Per the vendor, the number of Codex messages depends on the model, task size and complexity, and whether work runs locally or in the cloud. Plus and Standard Business get local-message estimates per five-hour period, given as a range per model; these are "not fixed message limits", and the usage dashboard shows current limits and reset times. Cloud tasks may use more of the allowance than local messages. Local messages and cloud chats share the plan's usage allowance, and weekly limits may also apply. Pro plans currently have no five-hour limit. ChatGPT Work shares the same pricing, credits and usage limits as Codex. Fast mode uses included usage at 2.5x (2x on purchased credits). Ultrafast uses 8x (6x on credits). Image generation uses limits about 3-5x faster. After reaching limits, Plus/Pro users and flexible-pricing workspaces can buy credits, which are billed per 1M tokens per model. Enterprise/Edu with flexible pricing have no fixed rate limits. Google AI (Gemini, Antigravity): In the Gemini app, Google meters compute used instead of daily prompts. Usage limits 'factor in the complexity of your prompt, the features you use, and the length of your chat. Your limit refreshes every 5 hours until you reach your weekly limit.' If you hit the cap on the biggest models, you are moved to smaller models. Media generation, Deep Research, the Pro model and Extended thinking/Deep Think use up the limit faster, and so do higher effort levels. Tiers are multiples: Plus is 2x and Pro 4x the 'standard' (no-plan) limit, and Ultra is 5x or 20x Pro. No prompt counts are published. Google Antigravity (app, IDE extensions and CLI) has its own agent quota that 'correlate[s] with the amount of work done by the agent'. Pro and Ultra get a quota refreshed every five hours up to a weekly rate limit; users without Pro or Ultra get a quota refreshed weekly. No figures are published. Pro and Ultra subscribers can buy AI credits for overage in Antigravity and Flow, and the subscriptions page now says credits can also extend Gemini app limits. Since June 18, 2026, Gemini CLI has been replaced by Antigravity CLI for unpaid and Google One users, so the old Gemini CLI daily request counts no longer apply to these plans.
Do ChatGPT + Codex and Google AI (Gemini, Antigravity) have free plans?
ChatGPT + Codex Free is free: Codex, GPT-6 Luna at Standard speed in the desktop app, subject to rollout; no message figures published; Codex (ChatGPT pricing page), Limited; Image generation in Codex, Not available; Codex Cloud, Not included. Google AI (Gemini, Antigravity) Free is free: Gemini app usage, Standard limits; Antigravity, Basic weekly rate limits; 'A meaningful quota, refreshed weekly'; Antigravity CLI (replaced Gemini CLI), Gemini CLI replaced by Antigravity CLI on June 18, 2026; covered by the Antigravity weekly quota, no request count published; Storage, 15 GB.
When do ChatGPT + Codex and Google AI (Gemini, Antigravity) limits reset?
ChatGPT + Codex: five-hour window, week. Google AI (Gemini, Antigravity): five-hour window, week, month. The figure inside each window is listed per plan on this page, or marked "not published" where the vendor gives none.
Limit alerts
Never get hit by a surprise AI quota cap.
One email the moment Anthropic, OpenAI, Cursor, Copilot or Google quietly change a 5-hour rolling limit, a rate limit or a plan price. Every alert links the vendor page and the old and new value.