Create AI models without writing code.
| Use Case | Classification | Extraction | Generation |
|---|---|---|---|
| Example | Identify a field in a form such as "Email address" or "Adresse e-mail" as "email". | Convert an email message into structured data to fill a form. | Answer support questions that users ask about your product via email. |
| Input | Structured data | Unstructured content | Unstructured content |
| Output | Structured data | Structured data | Unstructured content |
| Architecture | Embedding | Finetuning | RAG |
| Vendors | OpenAI, Gemini, Mistral | OpenAI, Gemini, Mistral | OpenAI, Gemini, Mistral |
| Cost | Low | Medium | High |
Train AI models effortlessly within Google Sheets—no coding required. Simply prepare your dataset, train the AI, and view results instantly through a formula in your spreadsheet. Once satisfied with the performance, you can seamlessly integrate the trained model into your app using our API or the Genkit API. This allows you to create custom, domain-specific AI solutions directly from your data in Google Sheets, making advanced AI accessible to users with or without a technical background.
Switch between AI architectures to fit your needs: use Embedding for tasks like identifying fields in a form, or Finetuning to convert email messages into structured data. For more complex tasks like generating detailed answers to support questions, RAG (Retrieval-Augmented Generation) is ideal. With our addon, you can adjust architectures seamlessly without altering your app, even after it’s live.
Switch between AI models from various vendors with ease. Select the best model for your classification, extraction, or generation tasks. Optimize for both accuracy and cost by evaluating and switching models to fit your budget and performance needs. With our addon, integration is simple—just hit “publish” to switch vendors without rewriting any API.
