Chat
The chat at typno.ai/chat works like the assistants you know: a conversation on the right, your earlier chats on the left.
What it can do
- Search the web. With Web on, the assistant searches and reads pages when a question needs current information, and names its sources. Switch it off for questions that should stay inside the conversation.
- Read documents. Add a PDF, Word document, Excel file (.xlsx), presentation or text file. The text is read out and given to the model. A scanned PDF has no text; add its pages as images instead.
- Look at images. Add or paste an image and ask about it.
- Make images. Ask for an image and the assistant generates one.
Add up to six files per message by clicking the paperclip, dragging them in or pasting.
Models
Pick a model at the top; the price per million tokens is shown next to each and your choice is remembered. Three models come first:
- Affordable: selected by default and fine for everyday questions.
- Recommended: the best choice for most questions.
- Most powerful: for hard work. It costs the most and reads no images.
Effort
For models that support it you choose next to the input how hard the model thinks, from fast to maximum. More effort helps with hard questions, but costs more tokens and time.
Costs
The chat is billed like the API, per token, to your organisation's AI service. The top of a conversation shows what it has cost so far; the sidebar shows this month and all your chats together. A long conversation costs more per question, because earlier messages are sent along each time. Start a new chat for a new subject.
Access
You sign in with your My Typno account. The chat works when your organisation has an active AI service; order one in My Typno. Your chats are yours: colleagues in your organisation cannot see them.
See Data and privacy for what is stored.