Guest Bots: On-Demand AI Inside Any Chat
Telegram’s new Guest Bots feature changes how users summon AI assistance. Instead of adding a bot as a contact or group member, you can now mention a Telegram AI bot by its @username directly inside any private or group conversation. The bot receives only the message where it was tagged and the related replies, then posts its response in the same thread. That design brings AI search, fact-checking, translation, image generation and other assistant tasks into existing chats without exposing full histories or participant lists. For everyday users, Guest Bots make Telegram guest bots feel more like context-aware helpers than separate apps. For developers, they create a low-friction way to offer AI tools that integrate into ongoing discussions, from customer support and team coordination to personal planning and study groups.
Multi-Bot Workflows and Streaming Responses
Beyond single assistants, Telegram is pushing toward full bot automation workflows. Bots can now talk to other bots, enabling chained actions such as one bot summarising a conversation while another schedules tasks or posts updates to a channel. This bot-to-bot communication turns Telegram AI bots into components of larger automation pipelines that operate entirely inside chat. At the same time, streaming responses let bots display text as it’s being generated, mimicking the typing flow of modern AI assistants instead of sending only final messages. Together, these chat automation features make Telegram a platform for real-time, multi-step automations that feel conversational rather than scripted. Teams, creators and power users can orchestrate complex processes—from content drafting to moderation—by simply mentioning the right bots in the right order.
Profile Chat Automation and Custom AI Writing Styles
Telegram’s new profile-level Chat Automation turns bots into personal representatives. Users can link a bot to their profile and let it respond on their behalf, with fine-grained controls over which conversations it may access. You can exclude specific contacts, limit automation to new chats, or use it only in certain contexts, making AI delegation safer and more predictable. For creators and organisations, custom AI Styles extend this automation to content creation. These styles are prompt-based presets in the text editor that lock in a tone or voice—such as formal updates, playful memes or branded copy. They can be shared via preview links so teams and communities maintain consistent messaging. Combined, profile automation and AI Styles transform Telegram from a simple messenger into an environment where personal and organisational personas can be partly run by configurable AI agents.
AI Sticker Search and Smarter Admin Controls
Telegram’s sticker ecosystem now benefits from AI-powered search across more than 100 million emoji and stickers in 36 languages. Using custom models on the Cocoon Network, Telegram scans and labels public sticker sets so users can find relevant reactions instantly by typing natural language, not just browsing packs. This makes expressive responses faster and connects years of user-generated media to practical bot automation workflows—for example, bots that suggest stickers based on conversation context. On the administration side, new tools strengthen group and channel governance. Poll statistics with vote graphs appear after 100 votes, while custom poll limits let admins target specific subscriber segments. Silent scheduled messages allow announcements without notifications, and reaction moderation lets admins remove reactions from particular users. These controls help communities maintain order and tone even as AI-driven bots and chat automation features become more active in public spaces.