Does GPT3.5 have usage limits? Can I use GPT-4?

Everyday chat is effectively unlimited, but remember this is a free-tier account: GPT-4-class models come with only a small trial allowance, and sharing the account drains it faster.

What the free tier can and can't do

• Basic chat: uses the free tier's default model. Writing, translation, rewriting, and general Q&A run without practical limits — you might only see slowdowns during OpenAI's peak hours.
• GPT-4-class models: OpenAI grants free accounts a small trial allowance that refreshes on a rolling window. When it runs out, you'll see a limit notice and the chat automatically falls back to the base model without interruption.
• Advanced features: file uploads, image generation, and data analysis are either heavily capped or unavailable on the free tier — don't rely on them.

How sharing affects your allowance

The trial allowance belongs to the whole account, not to each user. If others have already burned through the GPT-4-class messages, you may land straight on the base model when you sign in. That's normal for a shared free account, not a fault.

When you hit the limit notice

1. Read the notice — if it names a retry time, the allowance refreshes automatically then.
2. Carry on with the base model; it handles most text tasks well.
3. If GPT-4-class output is what you mainly need, upgrade to Shared GPT Plus (much higher limits) or Plus Top-Up (your own account, no sharing).

FAQ

• Why did I never get a single GPT-4 message? Other users on the account used the allowance first — try again soon after it refreshes.
• Replies are slow or say "try again later"? Free-tier traffic is deprioritized behind paid users at peak times; try an off-peak hour.