How can we help?
Everything you need to get started with Theorify — from your first workflow to advanced tips.
Getting started
Theorify is a visual workspace where you chat with multiple AI models (ChatGPT, Claude, Gemini, and more) on an infinite canvas. Instead of copy-pasting between separate AI tabs, you connect conversations so context flows automatically from one model to the next.
- Go to your Dashboard and click New project.
- You'll land on a blank canvas. Click + Add chat to drop your first chat node.
- Pick a model (e.g. Claude) and start chatting.
- Add a second node, pick a different model (e.g. GPT-5.2), and drag a connection from the first node to the second. Context carries over automatically.
Yes — the beta is completely free. You get full access to all supported models, so you don't need to add any API keys or payment info to get started.
How the canvas works
Each chat is a separate conversation with an AI — like Claude, ChatGPT, or Gemini. You can have as many chats on your canvas as you need. Think of them like sticky notes, but each one is a full AI conversation you can type into.
When you draw a connection from one chat to another, Theorify automatically passes the context of the first conversation into the second. The second chat 'knows' what happened in the first — no copy-pasting required.
A hand off passes your work from one chat to the next. Theorify packages the key points from your conversation and sends them to the connected chat. You can preview exactly what gets sent before you confirm.
Yes. A single chat can connect to multiple others. For example, you could research a topic with Gemini, then branch into two paths: one using Claude to write a report, another using ChatGPT to draft talking points.
AI models
Theorify supports models across multiple providers:
- OpenAI: GPT-5.2, o1, o3-mini
- Anthropic: Claude Opus 4.6, Claude Sonnet 4.6, Claude Haiku
- Google: Gemini 3.1 Pro, Gemini Flash
- Perplexity: Sonar (internet-connected search)
More models are added regularly. All models are available directly — no API keys or external accounts needed.
- Research & fact-finding: Perplexity Sonar (searches the internet) or Gemini (large context window)
- Writing & editing: Claude (nuanced, detailed prose)
- Reasoning & analysis: GPT-5.2 or o1/o3 (strong at structured logic)
- Quick tasks: Gemini Flash or DeepSeek (fast and free)
Memory
Theorify has three layers of memory: Global memory (your general preferences, like 'always be concise'), Project memory (context for a specific project), and Workstream memory (context for a specific workflow). Each layer automatically injects relevant context into your chats.
Open the Memory panel in your project (click the brain icon in the toolbar). You can add notes like 'I prefer bullet points' or 'My company name is Acme Corp'. Theorify will include this context in future conversations automatically.
Yes. Memory entries can be edited or removed at any time from the Memory panel. Changes take effect on the next message you send.
Billing & credits
During the private beta, Theorify is free — including complimentary AI credits. After the beta, we'll offer a paid plan (Pro) with transparent pricing. We'll give you plenty of notice before any changes.
Beta access doesn't expire while the beta is running. We'll give you plenty of notice before any changes.
Yes. Each project shows an estimated cost for the conversations within it. You can also see a usage breakdown in Settings. Costs are estimated based on published model pricing.
Troubleshooting
- Check the model provider's status page (OpenAI, Anthropic, Google) — outages happen.
- Try a different model — if Claude is down, switch to GPT-5.2 for now.
- Still stuck? Use the feedback button (bottom-right) to let us know.
Make sure the source chat has messages in it — empty conversations can't generate a summary to pass along. Also check that the connection arrow points from the source to the destination (not the reverse). If it still fails, try removing and re-drawing the connection.
Make sure you're logged into the same account you used to create the projects. If you signed up with Google, use the Google login option. If you used email, check that you confirmed your email address.
Contact
Can't find what you need? We're here to help:
- In-app feedback — Click the feedback button in the bottom-right corner of any page. We read every submission.
- Email — support@theorify.ai