Blockchain trading data, straight from the database
Read-only ClickHouse credentials for decoded trading data across every major Solana venue — Pump.fun, PumpSwap, Raydium, Meteora — plus HyperLiquid, Polymarket and Robinhood Chain. Connect with DBeaver, DataGrip, Python, or anything else that speaks ClickHouse. No API layer, no rate limits, no credits.
$200/month per chain · one-week free trialStart free trial · See a query
We also offer fast Solana nodes — gRPC, shreds and preconfirmations, from $200/month. Jump to nodes →
Pump.fun
Raydium
Meteora
HyperLiquid
Polymarket
UniswapA query, not a download
You get a hostname, a user and a password. That's the entire onboarding.
The ten most prolific Pump.fun creators of the last 30 days, and how much of each token's supply they bought into their own launch:
SELECT
creator,
count() AS launches,
round(avg(bundled_buys_count), 1) AS avg_bundled_buys,
round(avg(bundled_buys / if(mayhem_mode = 1, 2e13, 1e13)), 1) AS avg_bundle_pct_supply,
toTimeZone(max(block_time), 'UTC') AS last_launch_utc
FROM pumpfun_token_creation
WHERE block_time > now() - INTERVAL 30 DAY
GROUP BY creator
ORDER BY launches DESC
LIMIT 10| creator | launches | avg_bundled_buys | avg_bundle_pct_supply | last_launch_utc |
|---|---|---|---|---|
bwamJ…NfSXa | 17,229 | 1.3 | 21.5 | 2026-09-01 04:36 |
7naFF…AMidP | 5,264 | 4.0 | 30.2 | 2026-09-01 00:37 |
DmH2D…ygwcW | 4,520 | 1.0 | 15.1 | 2026-08-28 06:46 |
EBx24…VwrcD | 3,972 | 3.4 | 30.4 | 2026-09-01 03:11 |
AJNaS…fqvQ2 | 3,699 | 0.0 | 0.0 | 2026-08-31 23:07 |
2bRoa…ZzueC | 3,524 | 2.7 | 9.3 | 2026-09-01 05:32 |
FCUPh…BYhm9 | 3,478 | 1.1 | 17.9 | 2026-09-01 00:58 |
whamN…2oLbx | 3,269 | 3.1 | 31.1 | 2026-09-01 00:05 |
6ePbE…3tHHw | 3,229 | 3.6 | 25.7 | 2026-09-01 04:42 |
dtrzJ…MedXy | 3,020 | 2.0 | 24.0 | 2026-09-01 04:47 |
Run against the live database on 1 September 2026, 08:49 UTC. Addresses shortened for display.
The aggregation runs on our server. You get ten rows back, not ten million.
Setup takes about five minutes →
Why this is hard to get elsewhere
Depth
Pump.fun activity back to January 2024, across launches, bonding-curve trades, migrations and creator fees. Nothing sampled, nothing rolled up into hourly candles before it reaches you.
Resolution
Every swap carries the execution context you need to model it properly: pool reserves before and after, priority fees and compute units, the parent program that invoked the instruction, the fee payer, and slippage against the requested amount. Most providers hand you a price and a size.
Direct access
It's a ClickHouse database, not an API. Run a full table scan. Pull two years of swaps into a dataframe. JOIN across tables. Aggregate server-side and take back only the result. No rate limits, no row caps, no credits, no pagination. If a query gets heavy, we help you make it faster — we don't throttle you.
Coverage
Solana
Pump.fun · PumpSwap · Raydium · Meteora · since January 2024
Pump.fun and Pump.fun v2, PumpSwap, Raydium (AMM, CPMM, Launchpad), Meteora DLMM and Dynamic Bonding Curve. Plus token creation events, AMM migrations, Jito tips, SOL and SPL transfers, and block data.
HyperLiquid
Perpetual fills · order-book archives · since July 2025
Fills at the block level with millisecond fill times: price, size, side, position before the fill, realized PnL, fees, builder fees, liquidation detail and TWAP identifiers. Wallet and position views are computed at query time. Raw order-book archives support historical book reconstruction.
Polymarket
On-chain fills · market metadata · CLOB archives · since 2024
CTF Exchange fills with maker/taker side, fees and full EVM transaction context, joined to event and market metadata — questions, resolution sources, categories, liquidity and volume across AMM and CLOB. Raw snapshot and delta archives support book replay at a specific moment.
Robinhood Chain
Uniswap v3 and v4 · since May 2026
Decoded pools and swaps: signed balance deltas, square-root price and tick after each swap, in-range liquidity, pool fees and hooks, plus token metadata.
What people build on it
Train models, backtest strategies, analyze wallets, measure your own execution.
Research and modeling. Build features from decoded trade flow. Study trader and creator behavior, token launches, migrations and cross-venue activity.
Backtesting. Test signals against decoded trades and the order book as it actually was at that moment, not as it looks today.
Production. Dashboards, alerts, scheduled reports, notebooks and data pipelines on documented schemas that don't change under you.
More specifically:
- Research token launches, migrations, swaps, transfers, creator fees and trader behavior across the main Solana venues.
- Backtest signals against decoded trades and reconstructed order books, using the market state as it was at that moment.
- Analyze wallets, positions, fills, fees, slippage, pool state and protocol activity in plain ClickHouse SQL.
- Build reproducible notebooks, scheduled reports, dashboards and monitoring jobs on the same documented schemas.
Query examples → · Table reference →
We use this data ourselves
We built it for our own trading, then opened it up. Everything published on this site — every chart, every research table — is regenerated from the live database on each build. Nothing here is hand-drawn.
See our work on execution routing, token launches and wallet behavior in Research.
Then trade it
Trading has two phases. Research — you need the data. Execution — you need the nodes.
| Service | Price | Best for |
|---|---|---|
| Basic gRPC | $200/month | Real-time transaction feeds, account monitoring, dashboards |
| Very-fast gRPC | $700/month | Latency-sensitive production consumers |
| Ultra-low-latency gRPC | $1,200/month | Trading systems where ~1 ms matters |
| Dedicated custom gRPC | $3,300/month | Isolated node, custom configuration |
| Shreds | $500/month | Raw block propagation before RPC assembly |
| Preconfirmations | $1,600/month | Execution status before shreds and RPC commitment |
Same infrastructure the data comes from.
Getting access
$200/month per chain. Full access to that database, plus a direct support channel with the people who built the schemas.
One-week free trial. Same access as a paid account — no reduced tables, no row limits.
Start free trial · Talk to us on Telegram
Questions
What exactly do I get? A hostname, a username and a password for a read-only ClickHouse database. Use DBeaver, DataGrip, Python, curl, or anything else that speaks ClickHouse.
How fast do I get credentials? Message us on Telegram and we'll usually set you up right away. Through the form, within 24 hours — normally sooner.
Are there query limits? No rate limits, row caps, credits or pagination. You can scan whole tables.
What if my queries are too heavy? We'll reach out and help you optimize the query or the workflow. Our default is to make the workload fit, not to cut you off — and the people who wrote the schemas usually know the fast way to write it.
How fresh is the data? Tables update continuously, with a short delay behind the chain.
Can I export the data? Yes. It's a database — pull whatever you need, however you need it.
Do you cover a protocol that isn't listed? Ask. We add coverage regularly.