Chat with Content

Each text element (headline, body, related story, CTA) has its own chat panel so you can instruct the AI in natural language. The chat is context-aware of your source, rules, persona, language, and prior messages.

• Click the element you want to chat with
• Enter in any text you want
• Use the Quick prompts if you want inspiration

• Regular chat – you type free-form instructions.
• Quick chat – pre-made quick prompts for tone, length, communication models (AIDA, PAS, FAB), bullets, readability, and more. Users cannot create their own quick chats yet. Admins can request new ones to be added on the backend. Self-service creation is on the roadmap.

Manual edits change the text directly. Chat regenerates text using your rules and context. Magnity keeps these in sync so you can switch between them at any time without losing changes.

The text editor is where you can apply rich text, such as bold, underline, italic, bullets, links etc to the text Magnity has generated. Edit your text and submit, and it is added to the the interface.

Yes. Once you have applied formatting with the rich text editor, Magnity preserves bullets, bold, italics, and links even when you continue chatting.

When you create persona variants, each version inherits the chat history from the original base content. For translated versions you can open chat on the localized element and continue working in the target language.

Click the element you want to edit. The chat panel opens on the right. Type instructions like “Shorten to 100 words” or “Rewrite using AIDA without headlines.

Open the quick chat menu, pick a preset, and send it as is or tweak the text first. You can combine ideas, for example “Use AIDA, but do not include headlines.”

Yes. If the variant is in Korean and you type instructions in English, the output is returned in Korean for that variant.

Apply formatting first in the Text editor tab, then continue chatting. From that point the chat operates in HTML and preserves your formatting.

Yes. You can delete chat entries on an element to step back through previous states. This is useful if a series of edits takes the copy in the wrong direction.

Magnity detects your manual edits, updates the AI’s working version, and refreshes the chat context so your next instruction applies to the latest text.

• Personalization – persona variants start with the original chat history and remain persona-aware when you keep chatting.
• Translation – each language version supports chat; you can continue refining structure and tone in the target language.

• One clear intent per message, then iterate.
• Use quick chats for common edits, then fine-tune with a short follow-up. Ask for multiple options, then pick one, for example “Give me 5 headline alternatives, I’ll take number 3.”

Not yet from the front end. Ask your admin to have them added on the backend. Self-service quick chat creation is on the roadmap.