$42.5 billion.
That is the combined daily volume across the five centralized exchanges I pulled into this audit. Not Telegram's. Not TON's. The five CEXs that already move money between fiat and crypto every single day, while the conversation about Telegram becoming a global payments network keeps happening as if these rails do not already exist.
I want to talk about Telegram. I cannot talk about Telegram honestly without first talking about what it would actually have to displace.
Methodology
I started this piece trying to do something different. The plan was to map Telegram's published payments roadmap against its actual on-chain footprint and find the delta. What I ended up with was a problem of grounding: most of what circulates online about Telegram's payments ambitions is press positioning, not architecture. There is a real article waiting to be written about TON throughput, wallet activation, and the partnership economics — but it is not the article I can write responsibly with the data I have in front of me. So I pivoted.
The dataset I am working from covers five centralized exchanges: Binance, Bybit, Bitget, OKX, and MEXC. For each one I have founded year, headquarters jurisdiction, listed-pair count, daily spot-and-derivatives volume in USD, security score, proof-of-reserves audit date and status, full fee schedule, KYC posture at deposit, max leverage on futures, and the fiat onramps each of them currently runs in Brazil, India, and the EU.
This is not a Telegram audit. This is an audit of the rails any new "global payments network" would have to either replicate, bypass, or convince users to abandon. That distinction matters. Read the findings as a measurement of the wall, not of the climber.
Finding #1: The fee story is already over
Every single fiat onramp in my dataset charges the user 0% to deposit. Binance PIX in Brazil: 0%, instant. Bybit UPI in India: 0%, instant. Bitget PIX: 0%. OKX SEPA in Europe: 0%, settling in one to two days. MEXC PIX: 0%, instant. Binance UPI: 0%, instant. The pattern is universal at this tier of the market and it is universal across Latin America, India, and the eurozone.
So when the public framing of Telegram's payments push leans on "we will make sending crypto cheaper", I think the framing is already wrong. Cheaper than what? The retail user moving money into a major CEX in 2026 is not paying a deposit fee. They are paying — eventually, on the trade — somewhere between 0.02% and 0.10% in maker-taker. MEXC is the cheapest in the dataset at 0.00% maker and 0.02% taker. Binance, Bybit, and Bitget all sit at 0.10% on both sides. OKX is 0.08% maker and 0.10% taker.
That is not nothing for a high-volume trader. But for a user sending value as a payment — which is the explicit Telegram pitch — the trade fee is not where the friction lives. The friction lives in conversion, in custody, in jurisdictional acceptance. The "cheaper rails" story is the wrong door. If Telegram's value proposition is "we charge zero on deposits", every CEX in this dataset already charges zero on deposits. That race is run.
Finding #2: KYC at deposit is the real fork in the road
Four of the five exchanges in this dataset do not require KYC to deposit. Bybit, Bitget, OKX, and MEXC all let funds in before any identity verification. Only Binance requires KYC at deposit, and Binance is also the one with full or limited licenses across Dubai (VARA), France (AMF), and Italy (OAM). The correlation is not coincidence.
This is the actual structural dividing line in retail crypto right now, and it is also the dividing line that any "global payments network" runs into immediately. The thing that makes a payments network feel like a payments network is that you do not present a passport every time you receive money. The thing that makes a payments network legal is that someone, somewhere in the chain, does.
Telegram, as a messaging product, sits structurally closer to the first model than to the second. That is the appeal. It is also the regulatory cliff. Look at the four no-KYC-at-deposit exchanges in this dataset and where they are domiciled. Bybit holds full licenses in Cyprus (CySEC) and Dubai (VARA). Bitget holds full licenses in Lithuania (FCIS) and Poland (KNF). OKX has a provisional VARA license and a full Bahamas (SCB) license. MEXC is offshore Seychelles with no tier-2 license at all.
That is the menu. Frictionless deposits, but the licensing strategy is either creative jurisdictional shopping or no real license whatsoever. A payments layer that wants both global reach and fiat connectivity in the way Telegram's pitch implies has to pick. There is no fifth option hiding anywhere in the data.
Finding #3: The "verified" reserve label is older than it sounds
Every exchange in this dataset is listed as having a "verified" or "partial" proof-of-reserves status. The dates are where it gets interesting. Binance's last PoR audit is dated 2025-03-01. OKX, also 2025-03-01. Bybit, 2025-03-12. Bitget, 2025-02-20. MEXC is the outlier at 2024-12-10, and MEXC is also the one carrying a "partial" status rather than full verification.
Even at the best, you are looking at audit timestamps that are months old by the time the "verified" badge shows up next to the exchange name in any rating site. Proof of reserves is a snapshot. It is a Merkle tree of customer balances at a single moment, against an attestation of on-chain holdings at that same moment. It is not — and this matters — proof of liabilities. It does not tell you what the exchange owes elsewhere. It does not tell you whether the wallet was empty the day after the auditor walked out.
The reason this matters for the Telegram payments conversation is that any settlement layer that wants to behave like infrastructure has to solve the same problem these exchanges have not solved: continuous, liability-aware, real-time solvency. Otherwise the payments layer inherits the same accusation that broke FTX in November 2022 before the Alameda balance sheet leaked. "Verified" with a March date is not what infrastructure looks like. It is what marketing looks like.
Finding #4: The leverage gap tells you who these venues are actually for
The maximum leverage on futures across this dataset is wild and worth saying out loud. MEXC offers up to 200x. Binance and Bitget cap at 125x. Bybit and OKX at 100x. These are the same venues running the supposedly retail-friendly fiat onramps in Brazil and India. The same brand that takes your PIX deposit at zero fees will also let you go 125x long on a perpetual contract three taps later.
I want to be direct about what this means. Every one of these exchanges is structurally a derivatives venue with a payments interface bolted on for user acquisition. The PIX deposit, the UPI deposit, the SEPA wire — these are the funnel. The funnel feeds the futures order book. That is the business model. The reason the deposits are free is because the trading is not, and the reason the trading volumes look the way they do — Binance at $18.5 billion a day, Bybit at $9.2 billion, Bitget at $6.1 billion, OKX at $4.9 billion, MEXC at $3.8 billion — is leverage flow, not remittance flow.
A "global payments network" that wants to be a payments network and not a derivatives funnel has the exact opposite incentive structure. It needs the user to move money in, hold it for a moment, and move it back out, without ever ending up in a 100x perp. There is no business model in the dataset that looks like that. Telegram is, in theory, the company best positioned to build one, because it does not need futures revenue to subsidize the rails — it has a messaging business attached. Whether that thesis survives contact with the actual cost of running compliance, custody, and fiat connectivity at scale is the entire question. The numbers in this dataset do not answer it. They just tell you how much heavier the building is than any press release sounds.
| Exchange | Daily Volume (USD M) | Taker Fee | KYC at Deposit | PoR Last Audit | |---|---|---|---|---| | Binance | 18,500 | 0.10% | Required | 2025-03-01 | | Bybit | 9,200 | 0.10% | Not required | 2025-03-12 | | Bitget | 6,100 | 0.10% | Not required | 2025-02-20 | | OKX | 4,900 | 0.10% | Not required | 2025-03-01 | | MEXC | 3,800 | 0.02% | Not required | 2024-12-10 |
What This Does NOT Prove
I am not auditing Telegram. I want to be specific about the limit of what is on screen here. The dataset I am working from contains five CEXs and their published fee, license, and onramp data. It does not contain TON throughput numbers, TON wallet activation rates, the specific architecture Telegram has filed with any regulator, the exact language of partnership announcements, or the breakdown of stablecoin flows actually moving inside the Telegram wallet today. I cannot tell you what is shipped versus what is roadmap. I cannot tell you what the chargeback economics look like for a messaging-app payment, or how the custody is structured under the hood.
What this audit can tell you is what the entrenched competition looks like. It can tell you that "make payments free" is not a wedge, because deposits are already free at every venue I checked. It can tell you that the no-KYC-at-deposit posture every retail user secretly wants is currently held only by venues whose licensing strategy is either jurisdictional arbitrage or fully offshore. It can tell you that proof-of-reserves attestations are months old and do not include liabilities. It cannot tell you whether Telegram clears any of those bars. It can only tell you the height of the bars.
The Takeaway
If Telegram wants to be a global crypto payments network, the rails are not the moat. The licensing posture is. Whoever solves frictionless deposits with continuous, liability-aware solvency under real multi-jurisdiction licensing wins — and nothing in this dataset has done that yet.
FAQ
Are Telegram's payment fees actually lower than what these exchanges charge?
I cannot verify Telegram's published payment fees against the dataset I am working from, because the dataset is exchange-side. What I can verify is the claim it would have to beat. Every fiat onramp on Binance, Bybit, Bitget, OKX, and MEXC across Brazil, India, and Europe is 0% to deposit, with PIX and UPI settling instantly. The cheapest taker fee in the dataset is MEXC at 0.02%. So any "we are cheaper" pitch is not competing against expensive rails. It is competing against rails that are already free at the deposit step and very thin at the trade step. The fee story is not where Telegram wins.
Why does no-KYC-at-deposit matter for a payments network?
Because that is the user experience that makes a payments layer feel like a payments layer rather than a brokerage. Bybit, Bitget, OKX, and MEXC all let funds in without identity verification at the deposit stage. That is the closest existing analogue to how a messaging-app payments network would feel from the user side. The catch is the licensing trade-off. The four exchanges that allow this hold tier-2 licenses in Cyprus, Dubai, Lithuania, Poland, the Bahamas, or have no real license at all. There is no "frictionless plus fully licensed in twenty countries" entry anywhere in this dataset, and that is the regulatory ceiling any Telegram-style payments push would hit.
Is proof of reserves enough to trust an exchange — or a payments layer?
No, and the dataset is a good reminder why. Every exchange in the audit shows a "verified" or "partial" proof-of-reserves badge, with audit dates ranging from December 2024 to March 2025. Even setting aside that those snapshots are months stale by now, proof of reserves is not proof of solvency. It does not include liabilities — what the exchange owes to creditors, lenders, or affiliated entities. The collapse of FTX is the reference case here. An exchange could in theory produce a Merkle tree of customer balances on a Tuesday and be insolvent by Friday. Any payments layer that inherits this audit model inherits this exact problem.
Why are the exchanges with the lowest deposit fees also the ones offering the highest leverage?
Because the deposits are not the product. The leverage is. MEXC offers up to 200x. Binance and Bitget go to 125x. Bybit and OKX cap at 100x. The fiat onramps — PIX, UPI, SEPA, bank transfer — exist to bring users into the venue cheaply enough that some percentage of them end up in the futures order book, where the actual fees are paid and the actual margin lives. That is the business model that subsidizes the free deposits. A pure payments network does not have that subsidy. It has to make custody and compliance work without a derivatives engine carrying the cost, and that is a much harder spreadsheet.