How to enable AI with your own API
NexoPad AI works in BYOK mode: you choose the provider, paste your key, and control the model.
The extension does not force you into a single provider. You can choose from several services and save a different model for each one.
These functions are intended for Pro or Teams plans and require a valid API key from the chosen provider.
How to enable AI
Open AI settings from the side panel or the popup
The extension lets you choose the active provider before saving the key.
Paste the provider API key
The key is validated before being saved so you do not end up with a broken configuration.
Choose the model
You can use the suggested default model or type a custom name if your account supports another compatible model.
Go back to the side panel and test an action on real text
That confirms immediately that the provider, key, and model are working.
Providers and default models detected by the extension
| Provider | Default model | Useful note |
|---|---|---|
| OpenAI | gpt-4o-mini | You can change the model to another name compatible with your account. |
| Anthropic | claude-haiku-4-5-20251001 | Designed for fast responses with lower cost. |
| Gemini | gemini-2.0-flash | A light option for short tasks. |
| DeepSeek | deepseek-chat | Available as a configurable provider. |
| Groq | llama-3.3-70b-versatile | Useful if you already work with Groq infrastructure. |
| xAI | grok-3-mini | Also supports a custom model if your account offers it. |
| Mistral | mistral-small-latest | An additional provider supported by the extension. |
Privacy, errors, and availability
The extension is built with a BYOK approach: your API key is used to call the selected provider directly, not to send it to a NexoPad-owned backend.
If you see a provider unavailable error, review three things first: whether your plan enables AI, whether the key is still active, and whether the model is available for your region or account.