简体中文
EnglishWhat Data Sources and AI Models Does Promptly Support?
Updated
Promptly supports a wide range of data sources and all major model providers, so you can connect your own content to leading LLMs without writing code. According to the platform's own documentation, you can import data from Web URLs, sitemaps, PDFs, audio files, PowerPoint presentations, Google Drive, Notion exports, and YouTube links, then attach those sources to apps built on models from OpenAI, Cohere, Stability AI, Hugging Face, and others.
Data Sources You Can Connect
Promptly's no-code RAG (retrieval-augmented generation) pipeline lets you import existing data and connect it to LLM models. The supported sources listed on the site include:
| Data Source | Notes |
|---|---|
| Web URLs | Import website content directly by URL |
| Sitemaps | Use a sitemap.xml to pull in site-wide content |
| PDFs | Upload documents as data sources |
| Audio | Load audio files and use them as data sources in your apps |
| PowerPoint (PPTs) | Upload presentation files |
| Google Drive | Import from your Drive |
| Notion | Import Notion exports |
| YouTube | Add YouTube URLs to build AI applications on video content |
The workflow is: add your existing data from these sources, then attach the data sources to your app inside the app builder. This is what powers the "no code RAG pipelines" the platform advertises — your data grounds the model's responses rather than relying on the model's training alone.
AI Models and Providers
Promptly supports all major model providers. The site specifically names:
- OpenAI (GPT-powered models)
- Cohere
- Stability AI
- Hugging Face
- and more
Google PALM models are also noted as now available. You can mix and match these — Promptly supports model chaining, meaning you can chain models from different major providers within a single app. The platform also provides processors for higher-level functionality, so you don't have to wire every step manually.
How This Affects What You Can Build
Because data sources and models are decoupled, the same imported dataset can feed different model choices. For example, you could build a support chatbot grounded in your PDFs and Notion docs using an OpenAI model, then chain in another provider's model for a specific processing step — all without coding.
Beyond the app builder, Promptly offers:
- Embeddable widgets — copy-paste a code snippet to add a chatbot to your website, with customizable look and feel (including a bot logo).
- Slack and Discord integrations — build GPT-based Slack apps and Discord bots backed by your data, no coding required.
- APIs — run apps and endpoints from your own applications, and upload/manage data programmatically.
What to Check Before Committing
The page confirms broad data-source and model coverage but does not list every file type, size limit, or per-source sync behavior. If your data lives in a format not named above (for example, a proprietary database or a niche SaaS tool), verify support before building. Pricing details are referenced via a "Promptly Pricing" section and a "subscribe" signal, but no specific figures or free-tier limits appear in the available material — check the pricing page directly to confirm what your use case costs.
If your content is already in URLs, PDFs, Drive, Notion, or YouTube, and you want to run it on OpenAI, Cohere, Stability AI, or Hugging Face models without code, Promptly's supported set covers that path.