Plan tiers
AI Codes It has four plans — Free, Basic, Pro, and Max. All include the browser IDE, live dev preview, file explorer, and mobile version. Publishing requires Basic or above.
| Feature | Free | Basic | Pro | Max |
|---|---|---|---|---|
| Active projects | 3 | 10 | 20 | Unlimited |
| Published live sites | — | 3 | 10 | Unlimited |
| AI credits | Limited | Standard | More | Most |
| Custom domains | 0 | 1 | 3 | Unlimited |
| History backup | ✗ | ✓ | ✓ | ✓ |
| AI SuperPowers (Skills) | ✗ | ✗ | ✓ | ✓ |
| Inspect mode | ✗ | ✗ | ✓ | ✓ |
| Visual Edit mode | ✗ | ✗ | ✗ | ✓ |
| Terminal access | ✗ | ✗ | ✗ | ✓ |
| BYOK (own API key) | ✓ | ✓ | ✓ | ✓ |
| Git clone (GitHub, Gitea…) | ✗ | ✓ | ✓ | ✓ |
AI prompt limits
When using platform-provided models, prompts are counted against a rolling window. The limit resets automatically — you always see a live countdown above the chat input.
- A progress bar shows prompts used vs. remaining in the current window
- A countdown timer shows exactly when the window resets — down to the second
- Once the limit is reached, the send button is disabled until reset
- Prompt limits apply only to platform (admin-provided) models
Note: Using your own API key (BYOK) bypasses all prompt limits entirely.
BYOK — unlimited prompts
BYOK (Bring Your Own Key) lets you use your own API key from OpenAI, Anthropic, Google Gemini, Groq, Mistral, xAI, Cohere, HuggingFace, Fireworks, OpenRouter, and more. When a BYOK model is active, prompt limits don't apply and usage is billed directly by your provider.
Go to the Config tab inside the AI panel, enter your key, then select your provider and model. Your key is stored in your browser only — it never reaches our servers.