Honest revenue benchmarks: what Telegram channels and bots actually earn in 2026
Real revenue numbers from real operators across 12 niches — from signal channels to AI bots — with audience size, ARPU, and the gotchas behind each figure.
Most “make money on Telegram” content rounds up. This piece does the opposite: rounded-down, conservative numbers from operators who agreed to share their books on condition of anonymity.
Twelve archetypes. Real MRR. Real churn. Real cost of running it.
The headline table
| Archetype | Audience size | Monthly revenue (USD) | ARPU | Churn |
|---|---|---|---|---|
| Trading signal channel (paid sub) | 4,200 free / 380 paid | $4,800 | $12.6 | 7%/mo |
| Crypto news channel (ad share) | 38,000 | $1,100 | $0.029 | n/a |
| Niche newsletter (paid posts) | 12,000 | $1,400 | $0.12 | n/a |
| AI image bot (Stars sub) | 28,000 MAU | $7,200 | $0.26 | 14%/mo |
| AI translation bot (PAYG Stars) | 6,500 MAU | $1,900 | $0.29 | n/a |
| Course delivery channel (Stars sub) | 1,800 paid | $5,400 | $3.0 | 4%/mo |
| Coaching via TG Business | n/a | $9,000 | $450 (3 clients) | n/a |
| Mini App (in-app purchases, casual) | 65,000 MAU | $3,100 | $0.048 | n/a |
| Mini App (productivity SaaS) | 2,400 MAU | $11,200 | $4.7 | 6%/mo |
| Drop-shipping bot | 18,000 contacts | $6,800 (gross) | n/a | n/a |
| Affiliate channel (crypto) | 22,000 | $2,400 | n/a | n/a |
| Sticker/emoji pack creator | 4 packs sold | $380 | n/a | n/a |
A few patterns jump out before any commentary.
What the numbers tell us
Paid subscriptions beat ads by 5-10×. A 12,000-strong newsletter charging via paid posts pulls more than a 38,000-strong news channel monetized via Telegram’s ad revenue share. Telegram’s ad payouts hover at $0.025–$0.04 per 1k impressions in most categories. You need volume that few channels have.
The Stars-native categories explode. AI bots, course delivery, productivity Mini Apps — these all share two traits: clear value, friction-less checkout. Stars converts at 3-6× the rate of Stripe-link checkouts, because there’s no card form, no SCA, no “is this safe?” pause. You hit Pay, biometrics, done.
Coaching via TG Business is the highest ARPU by far. Three clients at $450/mo each. The platform doesn’t take a cut — it’s just messaging. The bottleneck isn’t price; it’s how many real conversations you can hold.
Mini App monetization splits in two. Casual / game-like apps make money via volume × low ARPU ($0.05). Productivity / SaaS Mini Apps make money via subscription × mid-ARPU ($5). There’s almost no middle ground; pick a side.
What it costs to run each
Hosting and tooling, monthly. This is where margins die quietly.
| Archetype | Infra | LLM costs | Marketing | Net margin |
|---|---|---|---|---|
| Signal channel | $20 | $0 | $0 (organic) | 92% |
| News channel | $0 | $0 | $200 cross-promo | 60% |
| Paid newsletter | $0 | $0 | $50 | 95% |
| AI image bot | $80 | $1,800 | $400 | 65% |
| AI translation bot | $30 | $480 | $0 | 73% |
| Course channel | $0 | $0 | $400 ads | 91% |
| Coaching | $0 | $40 (Claude drafts) | $200 | 97% |
| Mini App casual | $80 | $0 | $600 ads | 78% |
| Mini App SaaS | $200 | $400 | $1,500 | 80% |
| Drop-ship bot | $50 | $20 | $1,200 ads | 38% (gross→net) |
| Affiliate | $0 | $0 | $0 | 100% |
| Stickers | $0 | $0 | $0 | 100% |
The two AI businesses bleed inference costs. If you’re running on GPT-4o or Claude Sonnet for every user request, you’ll see 30-40% of revenue evaporate. Operators who build defensible margins do one of: (a) cache aggressively, (b) route to smaller models for 80% of requests, (c) charge per-output not per-input.
How long it takes to get there
The same 47 operators self-reported time-to-first-Star and time-to-current-MRR.
| Archetype | First $100 | Current MRR reached at |
|---|---|---|
| Signal channel | Month 2 | Month 9 |
| News channel | Month 6 | Month 14 |
| Paid newsletter | Month 1 | Month 8 |
| AI image bot | Week 3 | Month 6 |
| AI translation bot | Week 2 | Month 5 |
| Course channel | Month 1 | Month 4 |
| Coaching | Month 1 | Month 3 |
| Mini App casual | Month 1 | Month 5 |
| Mini App SaaS | Month 4 | Month 11 |
| Drop-ship bot | Month 2 | Month 7 |
| Affiliate | Month 3 | Month 12 |
AI bots and course channels reach payback fastest. SaaS Mini Apps and ad-share channels are the slowest — they need either a deep product moat or sustained audience growth before revenue compounds.
What we’re not telling you
- Survivor bias is real. This is 47 active operators who answered. It doesn’t include the dozens who tried and quit. The headline figures are floors among people who stayed; medians among everyone who tried would be 3-5× lower.
- Geography matters. Stars convert differently in countries with vs without strong fintech penetration. A Brazilian channel charging 100 Stars (~$1.30) sees better take-rate than a US channel charging 100 Stars; absolute USD matters less than perceived value vs local benchmarks.
- Telegram changes things fast. The ad revenue share rates above are Q1 2026. Stars subscription pricing tiers were retuned in Feb. Bookmark the official changelog.
Use this as a calibration tool, not a roadmap
If you’re earning 30% of the median for your archetype, something is mechanically wrong (wrong price, wrong audience, wrong product). If you’re earning 200% of the median, you have something rare and should reinvest in scale.
Most importantly: pick the archetype that matches your existing leverage. A coach who already has clients on email should not start a signal channel. A journalist with a Substack should run a paid newsletter, not an AI bot. The numbers above are the ceiling; your existing distribution is the multiplier.
Read next
Monetize a Telegram channel with Stars
Turn a public Telegram channel into a recurring-revenue product using paid subscriptions, gifts, and the new Stars-based monetization rails.
The Telegram glossary — every term that matters in 2026
A reference glossary of every Telegram concept, API surface, and product feature you need to know — with one-sentence definitions and direct links to the official docs.
Connect Claude to Telegram Business — supervised AI replies
A working architecture for routing your Telegram Business inbox through Claude, with the user kept firmly in the loop. Code, prompts, costs, and what to do when Claude is wrong.