Celestia's economy, in motion
Track the forces shaping data costs, validator yield and network capacity. Every number below comes from the dashboard's monitored sources.
The most boring chart in crypto – by design
This chart never spikes. No mint-day fee storms, no bidding wars – Celestia's fee market is engineered to be flat, so anyone building on it can plan costs months ahead. We track a variance index to prove the flatness stays flat.
Methodology
All formulas, data sources, and assumptions used in this dashboard.
Celestia $/MB = (total_fee_utia / 1e6) / (total_blobs_size / 1e6) x TIA_USD
Source: Celenium API /v1/head (total_fee, total_blobs_size) + Binance public ticker (TIA price)
ETH Blob $/MB = blob_base_fee_wei x 131072 / 1e18 / 0.121094 x ETH_USD
Source: Ethereum RPC eth_blobBaseFee + Binance public ticker (ETH price)
ETH Calldata $/MB = 16 x 1e6 x gas_price_gwei x 1e-9 x ETH_USD
Assumes 30 gwei gas price. 16 gas per non-zero calldata byte.
Inflation rate = annual_provisions / total_supply x 100
Staking APR = (inflation x (1 - community_tax)) / bonded_ratio
Community tax: 2% (from /cosmos/distribution/v1beta1/params)
Fee coverage = annualized_fees / annual_provisions × 100 (fees are paid to validators and delegators, not burned)
Fees annualized from total_fee / chain_age_seconds × seconds_per_year
Source: Celestia LCD (/cosmos/mint, /cosmos/distribution, /cosmos/staking)
$/MB = total_fee_usd / total_blobs_size_MB (from Celenium head + Binance TIA ticker)
Fee coverage = annualized_fees / annual_provisions × 100 (fees are paid to validators and delegators, not burned)
Annualized fees = total_fee / chain_age_seconds × seconds_per_year
Source: Celestia LCD (/cosmos/mint, /cosmos/distribution), Celenium /v1/head
Block load variance = standard deviation of avg_block_load_pct series (historical snapshots)
Lower variance indicates more predictable and stable fee behavior. Celestia's fee market separates blob fees from execution, leading to less volatility than monolithic chains.
Source: eco.json avg_block_load_pct series
Monetary & staking economics
Plain version: how many TIA exist, how many new ones are printed per year, and how much is locked up securing the network. These three numbers anchor everything else on this page.
Inflation & staking returns
Computed inflation rate and estimated staking APR derived from annual provisions, total supply, and bonded ratio. Community pool growth over time reveals protocol accumulation velocity.
2% of every block reward accrues to a treasury that TIA holders vote to spend on ecosystem grants – it currently holds ~3.45M TIA and grows every block.
Inflation schedule & projections
Celestia's inflation starts at 8% annually and decreases by 10% each year until reaching the 1.5% floor. This directly impacts staking returns and token dilution.
Annual inflation rate
Decreasing 10% per year from 8% until 1.5% floor
Projected TIA supply
Token supply forecast over the next 10 years
Key takeaway: As inflation decreases, staking APR will compress. Validators with lower commissions will become increasingly important for maximizing delegator returns. The network reaches its inflation floor of 1.5% approximately in –.
Who pays for security?
Fees on Celestia are not burned – they go to validators and delegators on top of newly minted TIA. Comparing the two shows how much of the security budget users fund today, and how much inflation still covers.
Celestia's issuance follows a published schedule, and the community has twice voted to steepen it. Since genesis, dilution is down ~71% – a network deliberately getting cheaper to run.
How stake – and support – is distributed
Who actually holds the network up? The top-5 validators carry about a quarter of bonded stake, the long tail carries the rest – and the Foundation Delegation Programme can support up to 35 of the 100 active validator slots. Selection considers uptime, infrastructure, ecosystem contributions and geographic resilience; Fibre support is required once live. Official framework · updated 1 Oct 2026.
Cohorts 1–7
How much of their own money is actually at risk?
Self-delegation is a validator staking its own TIA alongside its delegators – capital that gets slashed first if it misbehaves. It's the cleanest signal of skin in the game a validator can give, and it's public on-chain, updated with every lookup snapshot.
Custodial validators run by exchanges typically self-bond near zero – they secure the network with delegators' assets, not their own balance sheet. That's a business-model difference, not necessarily a red flag on its own.
How much reward money is just sitting there?
Every validator accrues two separate on-chain balances that sit unclaimed until someone withdraws them: rewards owed to its delegators, and commission owed to the validator itself. This isn't the revenue split between the two – that's fixed by each validator's commission rate, shown elsewhere on this page – it's simply how much value is currently parked on-chain instead of being claimed.
Validators tend to withdraw their own commission far more often than delegators claim rewards – which is why the commission balance stays small while the delegator-reward balance keeps building up.
What Fibre costs
Fibre is live on Mainnet. Every Fibre blob pays a fixed amount plus a fee for each 256 KiB chunk, in TIA, with no auction. Pick a blob size and a daily volume to see the bill.
Rollup DA budget
How much a rollup spends on data availability per month, posting the same data as standard blobs or through Fibre.
–
| Per batch | Per day | Per month | Horizon | |
|---|---|---|---|---|
| Fibre | – | – | – | – |
| Standard blobs | – | – | – | – |
Trend analysis
Key metrics plotted over all historical snapshots (one every ~2 hours since June 9). Note: the series has a collector maintenance gap Jun 30 – Jul 5.
What it shows: ~43% of all TIA stays locked securing the network, week after week – a stable, mature staking market.
What it shows: new TIA appears exactly on the published schedule – no surprises, no discretionary printing.
What it shows: usage stays far below capacity – the headroom that keeps fees flat.
What it shows: the active set holds steady at ~98–100 operators – seats are contested, the network stays full.
Correlations
Pairing metrics the collector actually tracks, over the same last-24-snapshot window as the rest of the dashboard. There's no stored price history here (only a live spot quote), so this isn't a price chart section – it's where load, backlog and stake data get checked against each other.
What it shows: whether block load moves with transaction count, or something else is driving it.
What it shows: whether pending transactions pile up because blocks are full, or for some other reason.
What it shows: whether a growing bonded ratio dilutes the top-10 validators' share, or they move together.
bot.fun: an agent economy on Celestia
bot.fun runs on Eden, an EVM rollup that pays its own gas in TIA and ships on the Celestia network (its own site even
reports through Celestia's analytics endpoint). Autonomous agents launch meme coins and trade them on bonding curves –
everything below is live from bot.fun's own public API (bot.fun/api/v1), including each
agent's and coin's own generated artwork.
What it shows: whether a busier hour on bot.fun is also a bigger-volume hour, or activity is just noisier.
Ranked by realized + unrealized profit, not by how much they traded.
Highest trading volume right now, regardless of price direction.
A real API, not a data dump
Every number on this dashboard comes from the same read-only API you can call yourself – free, no key, CORS open. It's a reflection of this site's own collector, not a wrapper around someone else's: network state, supply & staking, fee economics, a DA-cost calculator, the validator set, and the computed signals above, each as its own endpoint.
https://skyneticlabs.com/celestia/api/v1/
Live chain state – height, block load, timing, mempool.
/network
Block height, avg block time & load, peer count, unconfirmed txs, mempool bytes.
The monetary side – issuance, community pool, inflation rate.
/economics/supply
Supply, annual provisions, community pool, current inflation – all in utia.
Who's bonded, and how concentrated the stake is.
/economics/staking
Bonded ratio, active/inactive counts, top5/top10/top25 stake share.
What people actually pay, sampled from real transactions.
/economics/fees
Average & median tx fee, average cost per block, 24h fee-level series.
Price any payload against the live flat rate. The only endpoint that takes a real parameter.
/economics/da-cost?mb=24000
Cost per MB, cost per block, and calculator.estimated_cost_tia for your ?mb= size.
How much of the chain is actually blob traffic.
/economics/da-usage
PayForBlobs share of txs/fees, total blob bytes, gas used & wanted.
The full set, filterable and sortable server-side.
/validators?status=bonded&limit=10
Stake, commission, uptime, trust score. ?sort= any numeric field.
One operator's full on-chain profile.
/validators/<moniker-or-valoper>
Looks up by exact moniker or valoper address, case-insensitive.
The same computed signals shown live at the top of this dashboard.
/insights
Real 24h/7d deltas on key metrics, in plain English – template text over real numbers, not an LLM call.
Any tracked metric, as a time series.
/series/<key>?limit=200
One of ~25 keys (see /meta/status) – e.g. bonded_ratio_pct, p4b_tx_share_pct.
Is the data actually fresh, and what's it built from.
/meta/status
Feature flags, last snapshot time, full list of available series keys.
Machine-readable list of every endpoint above.
/
Same list this page renders from, as JSON – point a script at it.
What a specific namespace – a rollup, an appchain – actually costs.
/namespaces/<namespace_id>
Bytes posted, blob count, estimated cost at today's rate, benchmarked against current network conditions. Built for the SDK below.
A local-first Python SDK for rollups & apps – track your own namespace's DA cost over time, on your schedule. We supply the live numbers; you own the history.
curl https://skyneticlabs.com/celestia/api/v1/insights