Thread conversations in Mantle Chat
Reply to any message in chats or channels, just like Slack or Discord.
The best part: You can thread with AI models and your custom agents together with your teammates.
#ProductUpdate#AI#AIChat#TeamCollaboration#MantelChat#SaaS
Mantle Chat February Update is here
What's new:
• Slack-like threads in chats & channels
• Real-time sidebar sync (ElectricSQL + PGlite)
• Claude Opus 4.6
• Workspace deletion with content migration
• Smarter channel editing & AI controls
Full update: mantle.chat/blog/monthly-u…
Small UX details matter
When you delete a workspace, we show a "Move conversations" option but only if you actually have chats or channels to move.
Less noise. Smoother flow.
We added new notification system for @`mentions` in channels at Mantle Chat.
When someone mentions you, you'll know about it instantly.
How It Works
🔔 Bell icon in header: Shows your unread notification count
📨 Notification panel: Click to see all your @`mentions` in one place
🎯 Click to navigate: Click any notification to jump directly to that message
Mantle Chat January 2026 Update 🚀
• Notification system for @`mentions` in channels
• MCP integrations: AI interacts with Linear, Notion, GitHub, Stripe
• Direct integrations: Real-time updates in channels (Linear, Stripe)
• AI Agents now generally available
• 10 AI models on free tier
• New rich text editor
• Rebuilt emoji picker
Read more: mantle.chat/blog/mantle-ch…#AI#ProductUpdate#Productivity#Collaboration#MantleChat
Working on notifications for @MantleChat🔔
What we did:
• MVP scoping — what's essential vs nice-to-have?
• "Inbox" vs "Notifications"? MVP is just @mentions, but we're planning AI responses, reactions, invitations, and more
• Picking the notification sound 🔊
AI-Writing Tools in @MantleChat
✍️ Fix spelling + grammar instantly
🔄 Paraphrase your message (more professional or simpler)
📝 Summarize long text into a one-click short version
We've introduced a new URL structure that keeps your workspace context in @MantleChat:
Benefits
• Bookmarkable: Save links to specific workspaces
• Shareable: Share workspace-specific URLs with teammates
• Consistent navigation: Workspace context persists across the app
We've implemented provider-agnostic prompt caching across all AI providers in @MantleChat
Cached prompts skip redundant processing, resulting in faster response times — especially noticeable in longer conversations. This works transparently across Anthropic, OpenAI, Google, xAI, and DeepSeek models without any changes to how you use Mantle Chat.
December 2025 and early January 2026 were very productive for @MantleChat
We added 7-day free trial for Pro plan, created the first documentation, added new articles in our Help Center, shipped AI Assistants, expanded our model lineup with new AI models, introduced voice-to-text, updated channels: added message editing, deletion, assistant mentions, fixed some bugs and even more.
Read a full article in our Blog:
mantle.chat/blog/mantle-ch…
What are Channels in @MantleChat?
Channels in Mantle Chat are shared chat rooms for team discussions where you can bring AI into the conversation at any time.
You can organize channels by topic, team, project, or purpose — similar to channels in Slack or Discord.
You can call AI when you need it:
• Mention @ models: GPT, Claude, Gemini;
• Mention @ your custom assistant: EnglishPracticeAssistant, ProcrastinationCheck;
If you're interested, you can find full article in our Help Center:
mantlechat.featurebase.app/help/articles/…
This week, we'll dive deeper into channels in Mantle Chat—what they are, how to use them, and a quick demo.
Today, I’m going to start with a quick survey.
Tomorrow, I’ll explain what Channels in Mantle Chat are.
Today we squashed an important bug introduced not so long ago. Several OpenAI models (GPT-4o, 4o mini, 4.1 nano, 4.1 mini, 5 nano, 5 mini, o3-mini, o3, o4-mini) weren't returning responses on Mantle Chat. Due to some prompt caching work, we used options that only GPT-5.1+ models support, which caused malfunctions on older models. Now fixed and all models work as expected!
#BuildingInPublic day 9