The chat runs in your browser: open the page and start typing. No separate apps, extensions or setup required.
Top up your balance once and use every model — no separate accounts or payment details with each provider.
Instead of a separate subscription for each service — one balance, deducted per token used.
A workspace for tasks that need you rather than a bot: analyze a document, write a text, run the numbers, test a prompt before putting it into a flow.
Switch models right in the conversation: one for quick answers, another for deep analysis. Keep the ones you use most in favorites.
Upload a document or a screenshot and ask about its content. Files stay in your library — you can attach them to another conversation.
Switched on with a toggle in the chat: the model pulls fresh data and can open a page by link.
The same document search the AI agent uses in your bot: ask about your own price lists and policies right in the chat.
For complex tasks, switch on reasoning mode and pick its depth — you see the line of thought, not just the answer.
Generation runs on the server: close the tab or lose the connection — when you come back, the full reply is there.
When a conversation grows, compress the context with one click — the conversation continues, and token usage drops.
Replies render properly: code highlighting with copy, tables and math instead of raw text.
The platform has 40 models. Each block and each chat gets its own model: some tasks need the smartest one, others a fast and cheap one.
The AI chat and the AI agent in your bot share the same models, knowledge base and balance. That makes the chat a handy test bench: check the wording and the model, then put them into the bot's flow.
How the AI agent works in a bot →You pay for the chat in tokens only. No separate subscription for each model, and no builder plan is needed for the chat.
A separate prepaid balance: deducted per token actually used, with no subscription fee for each model.
Billed monthly and deducted from your account balance. The price depends on the embedding model you choose and the size of the base. Auto-renewal can be turned off.
The balance is prepaid: only what you actually use is deducted. When it runs out, top it up with any amount.
The logs show which model was used, how many tokens it took and what it cost, broken down by model. It's easy to estimate your spend on a small volume.
No. The chat runs in your browser: open the page and start typing. No separate apps, extensions or setup required — every model is available right after you sign up.
With your AI balance: top it up with the amount you need, and it is deducted as you go — per token used. No separate subscription for each model, and no accounts with each provider. You don't need a paid builder plan for the chat itself — that's only needed when you launch bots.
The catalog currently has 40 models: GPT, Claude, Gemini, Grok, DeepSeek, Qwen and more. You switch models right in the conversation; each one in the catalog has a description, supported data types and pricing.
Yes: attach documents and images to a message and ask about their content. Uploaded files stay in your library, so you can attach them again in another conversation. Images are understood by models with vision support.
Nothing is lost: generation runs on the server, not in your browser. You can close the tab or lose the connection — when you return to the chat, the full reply will be there.
The chat is a workspace for you: you ask the questions and get the answers. The AI agent is a block in the bot's flow that answers your customers in messengers without your involvement. They share the model catalog, knowledge base and balance, so the chat is a handy place to test a prompt before putting it into a flow.
Signing up takes a minute: every model is available right away, and you pay only for what you use.