How to read TON on-chain metrics: an analyst's guide

TON on-chain metrics look familiar to anyone trained on Ethereum or Solana, but the semantics often differ. Sharding, asynchronous messages, a custom virtual machine and the tight Telegram coupling all create traps in interpretation. This guide is about which numbers actually matter, where to find them, and how not to fool yourself.

What an on-chain metric means in the TON context

Any metric is an aggregation of blockchain events. On Ethereum, an “event” is a transaction or an emitted log. On TON it’s more complex: a transaction here is a smart contract processing a message. A single user “transfer” produces a chain of 2-4 transactions (external message from the wallet → get-method call → internal message to the recipient → reply).

That changes how you read “transactions per day”. Tonstat.com shows ~5.5M transactions per day as of May 2026 — but that is not 5.5M user actions. The real action count is roughly 2-3x lower.

!

Trap #1 — TPS

Comparing “transactions per second” between TON and Ethereum head-on is wrong. TON’s number is inflated by internal contract-to-contract messages. For a fair comparison, count “economically meaningful transactions” — Toncoin transfers, jetton operations, NFT actions.

Baseline metric stack

Most analysts start with these five:

1. Active addresses (DAU/MAU) — unique addresses that sent at least one transaction in the period.
2. Transactions per day — total processed transactions.
3. Network fees — TON burned/paid to validators in fees.
4. TVL — total value locked in the network’s DeFi protocols.
5. Stablecoin supply — USDT/USDC/USDe issued on the network.

These five give a first read on network health. From there you drill into categories (DEX volumes, NFTs, jetton activity, staking).

Where to find data

Baseline — explorers and aggregators

For most jobs, those five sources are enough. The May 2026 baseline picture from tonstat.com:

• 181M total addresses;
• 53M activated on-chain wallets;
• 171,116 daily active wallets;
• 1.53M monthly active wallets;
• 5.59M transactions per day;
• 50,327 new activations per day;
• 370 validators, 845M TON staked.

Advanced — RPC and indexers

When the baseline dashboards aren’t enough, you read data directly via the TON HTTP API (toncenter.com) or run an indexer (ton-indexer from ton-blockchain). That lets you build custom slices unavailable in off-the-shelf services. For example, counting active addresses specifically in the mini-app category, or filtering jetton transfers by a particular issuer.

Dune has ready-made TON SQL tables, but coverage is thinner than Ethereum’s — many important dashboards still have to be hand-built.

Reading “active addresses” on TON

This is the most misleading metric. On Ethereum, DAU tracks user count fairly steadily. On TON you have three parallel layers:

1. Telegram MAU — around 1B (Pavel Durov, 2025). Off-chain.
2. Activated TON wallets — 53M (tonstat.com, May 2026). The wallet smart contract has been deployed but isn’t necessarily used.
3. Daily active addresses — 170,000-500,000 depending on the month. The “real users of the network”.

Between these layers there’s a huge funnel drop. Most in-Telegram Wallet users never activate an on-chain wallet (the in-Telegram Wallet is custodial — it doesn’t write to the chain at sign-up). Most on-chain wallet owners are inactive on a monthly horizon.

i

When DAU spikes

In December 2024 TON DAU reached 880,000 at the peak of Hamster Kombat. By December 2025 it had fallen to ~170-200k — that’s not a “network failure” but a return to baseline after a speculative airdrop cycle. Reading TON DAU growth and decline always requires accounting for airdrop waves.

Reading transactions

5.59M transactions per day sounds impressive — TON sits in the top 3 chains by transaction count. But “economically meaningful” transactions land 2-3x lower. Useful breakdown categories:

• Toncoin transfers — simple P2P transfer of the native token;
• Jetton transfers — operations with USDT, USDC, NOT and other Jetton-standard tokens;
• NFT actions — mint, transfer, sale;
• DEX swaps — STON.fi and DeDust interactions;
• System — wallet deployments, contract upgrades, gas computations.

Tonstat.com and tonviewer.com show this breakdown. The most telling growth signal is the share of jetton operations in total volume: the higher it is, the more “real” traffic the network has compared with one-off mini-app interactions.

Network fees

Tonstat.com shows ~4,289 TON per day in fees as of May 2026 (at $2.55, that’s ~$11,000/day). Compare to Ethereum, where fees usually run $1-3M per day. The two-orders-of-magnitude gap reflects TON’s low fees, not low activity.

Meanwhile TON’s inflation rate is 0.899% per year (source: tonstat.com) — new TON issued to validators as rewards. Part of fees is burned, partially offsetting issuance.

DeFi metrics

TON DeFi has its own quirks — covered in detail in the TVL and DEX-volume guide. Just the baseline figures for May 2026:

• Ecosystem TVL: around $300-400M (DefiLlama). Peak: ~$650M in mid-2024.
• Daily DEX volumes: $15-40M depending on market conditions.
• STON.fi share: roughly 60-70% of DEX volume; DeDust 15-25%.
• Staking: ~96M TON in liquid staking (Tonstakers, Hipo, bemo).

Comparing TON to other chains

Direct comparisons of “TPS”, “transactions/day” or “addresses” mislead. Better approaches:

• Per user: TVL/DAU, fees/DAU, 30d retention.
• By protocol revenue: STON.fi vs Uniswap by revenue (Token Terminal).
• Activity density: percentage of activated addresses that are active — TON’s is materially lower than Solana’s or Base’s.

i

A useful exercise

Take three networks — TON, Base, Sui — and assemble a five-metric table for one date. You’ll see that raw numbers say little; what matters is dynamics, retention, protocol revenue and the stablecoin share of TVL.

Common mistakes by junior analysts

1. Adding Telegram MAU to on-chain DAU. Two different metrics — never sum them.
2. Comparing pure TPS. TON is inflated by internal messages.
3. Reading TVL without staking context. Tonstakers/Hipo/bemo are sometimes counted in TVL, sometimes not — depends on DefiLlama’s methodology. Always check.
4. Ignoring airdrop waves. Any DAU/TVL spike following a major airdrop project (NOT, DOGS, MAJOR) is speculative, not organic.
5. Trusting social-media numbers. Twitter analysts often post figures without methodology. Cross-check with tonstat.com and DefiLlama.

Further reading

• TON DeFi metrics in detail — TVL, volume, active users.
• TON vs BTC and ETH correlation — historical analysis.
• Foundational TON architecture — complete guide.

Sources

• tonstat.com — main TON on-chain aggregator.
• tonscan.org — block explorer.
• tonviewer.com — alternative explorer with extended analytics.
• DefiLlama / TON — TVL and protocols.
• Token Terminal — TON — DAU and revenue.
• Gate.io research — On-chain TON 2025 — methodology overview.