Mira and Shieldy don't compete — they're two narrow, single-purpose tools that live on different Telegram surfaces. Shieldy (@shieldy_bot, open source on GitHub since 2018, MIT licensed) does one thing well for Telegram groups: captcha-on-join verification and anti-spam. New members prove they're human before they can post, blocking the spam-bot wave. Mira (launched 2026) does one thing well for Telegram channels: fetches RSS, YouTube, Reddit, and Google News, AI-scores for relevance, and publishes on schedule. If you only run a group, you want Shieldy. If you run a channel, you want Mira. Channel-plus-linked-group is the common pairing — both run side-by-side without conflict.
Searches for "mira vs shieldy" usually come from admins who like single-purpose tools and are wondering whether the bot they trust for entrance filtering can also handle content publishing — or whether a content bot could replace their spam defence. This post is the honest disambiguation, written from Mira's side. If anything is wrong, tell us in Telegram and we'll fix it.
Two narrow tools, two different Telegram surfaces
Telegram has two ways to publish or talk to many people at once:
- Channels — one-way broadcast. Only admins post. Subscribers read and react with emojis but can't send messages. Used for news, blogs, brand updates, curated feeds.
- Groups (or supergroups) — two-way conversation. Every member can post. Used for community chat, support, internal coordination.
The bots that live on each surface solve different problems. Shieldy lives in groups, doing one specific job. Mira lives in channels, doing one specific job. If you're not sure which surface you have, see our channel vs group breakdown.
What Shieldy actually does
Shieldy is the result of a focused product decision: do entrance verification really well, and skip everything else. The functionality (open source on github.com/1inch/shieldy) covers:
- CAPTCHA on join — multiple verification methods that new members must pass before they can post in your group. This is Shieldy's core feature
- Anti-spam — automatic blocks for known spam patterns and known spam accounts
- Restrict before verify — until a new member completes the captcha, their posting permissions are restricted, so spam-bot accounts can't blast links the moment they join
- Auto-removal of unverified joins — accounts that don't solve the captcha within a time window get kicked automatically
That's deliberately the whole list. Shieldy doesn't try to be a full moderation suite — no analytics dashboards, no welcome buttons, no notes-and-filters system, no federations. It's a single-purpose defence layer.
The bot is open source under MIT license, written in TypeScript, with a Premium tier (monthly, yearly, lifetime via Stripe) that unlocks more advanced verification options for larger groups. Self-hosting is also supported for teams that need full control.
What Mira actually does
Mira's home turf is the automation of Telegram channels (one-way broadcast). It connects to your sources — RSS feeds, YouTube channels, Reddit subreddits, Google News searches, public web pages, and public Telegram channels — scores each item with AI against your channel's voice, drops cross-source duplicates, and publishes on the schedule you set. Drafts arrive in your Telegram chat for one-tap approval, or auto-publish takes over on Pro. More on the flow in how it works.
Mira does no group moderation. No captchas, no entrance verification, no anti-spam. Channels don't have member posts, so there's nothing to verify or filter. These concepts simply don't apply.
Side-by-side: jobs and capabilities
| Capability | Mira | Shieldy |
|---|---|---|
| Telegram surface | Channels (one-way broadcast) | Groups / supergroups (two-way chat) |
| Primary job | Auto-publish curated content | Block spam-bots from joining the group |
| Fetch content from RSS / YouTube / Reddit | Yes | No |
| AI relevance scoring of articles | Yes — AI-curated to your channel's voice | No (out of scope) |
| Scheduled publishing with even pacing | Yes | No (groups don't need scheduling) |
| Cross-source duplicate filtering | Yes — semantic deduplication | No |
| CAPTCHA on join | No (channels don't have join captchas) | Yes — multiple methods, core feature |
| Restrict new members until verified | No | Yes |
| Auto-removal of unverified joins | No | Yes |
| Notes / filters / federations | No | No (out of Shieldy's scope too) |
| Analytics dashboard | No | No |
| Open source | No (closed-source SaaS) | Yes — MIT, github.com/1inch/shieldy |
| Self-hostable | No | Yes |
| Pricing | Free + Pro 1000 ⭐/mo (~$13) | Free + Premium (monthly / yearly / lifetime via Stripe) |
The two bots share zero capabilities. Both are narrow by design — Shieldy on the group-entrance problem, Mira on the channel-content problem.
The 30-second decision
Choose Shieldy if:
- You run a group or supergroup (members can post)
- You're tired of spam-bot accounts blasting links the moment they join
- You want captcha-on-join and entrance filtering, nothing else
- You prefer open source tools you can audit or self-host
- You already have a moderation layer and just need a defence-at-the-door piece
Choose Mira if:
- You run a Telegram channel (only admins post)
- You want to auto-publish content from RSS, YouTube, Reddit, Google News
- You need AI-curated relevance to filter noise
- You publish several posts per day on a schedule
- You want cross-source duplicate filtering
- You manage everything inside Telegram — no separate dashboard
Three real scenarios
Scenario 1: A standalone community group hit by spam-bots
You run a 10,000-member developer group. Every day, fresh accounts join and immediately post NFT scam links. Your admins ban them, but more keep coming. You don't want a heavy moderation suite — you just want the door closed.
Pick Shieldy. Captcha-on-join solves exactly this problem. Spam-bot accounts can't pass the verification, so they never post their first link. Mira is irrelevant here — it doesn't operate in groups at all.
Scenario 2: A standalone publishing channel
You run a Telegram channel about indie game development for 3,000 subscribers. You post 4–6 curated articles per day from a dozen RSS feeds and two YouTube channels. There's no group attached — readers just react with emojis.
Pick Mira. RSS / YouTube fetching, AI scoring, scheduled publishing — exactly Mira's job. Shieldy doesn't fetch content and doesn't work on channels.
Scenario 3: A channel with a linked discussion group (the common case)
You run a media project: a public Telegram channel where you publish curated news, plus a linked discussion group where readers debate the posts. Channel: 22K subscribers. Group: 1.8K active members. The channel currently eats two hours of your morning to curate, and your group has a slow but steady spam-bot trickle.
Run both. Mira automates the channel — RSS, YouTube, and Reddit feeds get AI-scored, duplicates filtered, drafts dropped in your chat for one-tap approval, or auto-published on schedule when you trust the settings. Shieldy guards the discussion group — captcha-on-join stops the spam-bot accounts before they post. Two narrow tools, two problems solved, zero overlap.
Frequently asked questions
What's the main difference between Mira and Shieldy?
Shieldy is a narrow anti-spam tool for Telegram groups focused on captcha-on-join verification — it stops spam-bots from joining and posting. Mira is a narrow content-automation tool for Telegram channels — fetching from RSS, YouTube, Reddit, AI-scoring, and scheduled publishing. They operate on different Telegram surfaces and don't compete.
Can Shieldy post content from RSS or YouTube to my channel?
No. Shieldy does one thing — captcha-on-join entrance filtering — and intentionally nothing else. It doesn't fetch external content, schedule posts, or curate articles. For channel content automation you need a different tool, like Mira.
Can Mira protect my group from spam-bots and verify members on join?
No. Mira is built for one-way channel publishing. It does not handle group moderation, captchas, anti-spam, or member verification. If you run a community group, use Shieldy for entrance filtering, or look at Combot or Rose for fuller moderation.
Does Shieldy work on Telegram channels?
Shieldy is built for groups. Its core feature — captcha-on-join — only applies to surfaces where users join and post, which channels are not. Shieldy has no role in channel content publishing.
Should I use both Shieldy and Mira?
Yes, if you run a channel with a linked discussion group — the common configuration for media projects. Mira automates content into the channel; Shieldy stops spam-bots from joining the discussion group. They don't conflict because they manage different chat IDs.
Is Shieldy free? Is it open source?
Yes to both. Shieldy is open source under the MIT license — codebase at github.com/1inch/shieldy. Free standard use, plus a Premium tier via Stripe (monthly, yearly, lifetime) for advanced verification features. Mira is closed-source SaaS, free to start, with a Pro tier at 1000 Stars/month (≈$13) that unlocks AI scoring at scale, auto-publish, and non-RSS sources.
I already have Rose or Combot — do I still need Shieldy?
Probably not for most cases. Rose and Combot both include captcha-on-join as part of broader moderation suites. Shieldy is for admins who specifically want a single-purpose tool — minimal surface area, open source, no extra features to misconfigure. If your stack is already heavier and working well, no reason to add Shieldy.
Picking between channel-automation tools
Shieldy isn't really a Mira alternative — it's a different category and a different surface. If you're comparing tools that actually do channel content automation, see Mira vs Junction Bot (web sources vs Telegram sources) and Mira vs IFTTT (single-purpose AI tool vs general workflow builder). For the parallel "group moderation, broader scope" comparisons, see Mira vs Combot and Mira vs Rose.
If you have a channel — Mira is the narrow tool for that job.
Three days of Pro free, no credit card. Shieldy guards your group's door; Mira fills your channel. Two narrow tools, two clear jobs.
Open in Telegram