Gemini AI Free Tier Models

gemini-3-flash-preview: 20 calls / day

gemini-3.1-flash-lite: 500 calls / day

gemini-3.5-flash: 20 calls / day

gemini-3.5-flash-lite: 500 calls / day

gemini-3.6-flash: 20 calls / day

These 5 models apply to both old and new API Keys. For older API Keys, there are 2 additional models:

gemini-2.5-flash: 20 calls / day

gemini-2.5-flash-lite: 20 calls / day

API Keys created before July 18, 2026, will get a total of 1,100 Calls / Day across 7 API models. Meanwhile, API Keys created on or after July 18, 2026, will get a total of 1,060 Calls / Day across 5 models.

The Lazy Way to Write API Calls

If you don't feel like keeping up with API Key updates, you can simply write these 2 models instead:

gemini-flash-latest

gemini-flash-lite-latest

When newer models are updated in the future, you won't need to edit the model names anymore. However, you will obviously only be using 2 out of the available 7 models (if 7 are available).

For now, as I'm writing this, gemini-flash-latest will point to/use gemini-3.6-flash, while gemini-flash-lite-latest will point to/use gemini-3.5-flash-lite.

If you still get a free tier allowance in the future, these model references will automatically follow the newest model version. If you are using a paid API plan, you can use gemini-pro-latest—though it seems very few people use the Pro API, since Indonesians / Indonesian citizens love free stuff! Lol.