Hyperliquid factors and forecasts, for your model.

Proprietary, model-ready factors and ML forecasts from the Hyperliquid DEX — a funding & volatility forecast and wallet-behaviour (toxicity) scores — built on clean, order-flow-enriched OHLCV, funding and positioning data. Stack the pieces into your own model. Parquet over a REST API, the whole universe, from October 2025.

Get a free API key How it works

No card required — the free tier is order-flow-enriched minute OHLCV plus a taste of the funding & volatility forecast, for BTC, ETH, SOL & HYPE.

curl https://tesseralytics.dev/v1/datasets \
  -H "Authorization: Bearer $TESSERA_API_KEY"

BTC, minute by minute

Daily close from gold_ohlcv_1m (Mar–May 2026) and the cumulative volume delta beneath it — the same order-flow-enriched Parquet the free tier serves, here charted straight from the data.

The forecast layer

A funding & volatility forecast and wallet-behaviour (toxicity) scores — orthogonal, supervised-learning factors you stack into your own model, not standalone signals. Validated out-of-sample on coins and wallets the models never saw.

The raw tiles, clean and checked

Order-flow-enriched OHLCV, funding and positioning — raw Hyperliquid fills deduped, aggregated and quality-checked end to end, so every candle and funding print is consistent before it reaches you.

Unlimited Parquet downloads

Every partition — raw data or factor — is a plain Parquet file. Drop it straight into Polars, pandas or DuckDB. Pull as much as you want; there's no per-GB fee.

Browse the full data catalog →

AI native

Skip the download scripts. Our hosted Model Context Protocol server exposes every gold dataset to Claude — or any MCP client — as a set of tools, authed with the same API key. The model discovers the catalog, reads the data dictionary, and pulls the rows itself. Point your client at it in one command:

claude mcp add --transport http tessera https://tesseralytics.dev/mcp \
  --header "Authorization: Bearer $TESSERA_API_KEY"

Then just ask — "what's the latest funding-innovation forecast for BTC?" — and it reasons over the data. See setup and example prompts →

How it works

Sign up with Google or your email — no card required.
Create an API key in your account — it's shown once, so save it as $TESSERA_API_KEY.
Install the Python SDK:
```
pip install "tessera-api[duckdb]"
```

Query straight into DuckDB SQL — one call spans as many coins and months as you like (MonthSpan fetches, range-reads and unions the partitions for you):

import tessera

client = tessera.TesseraClient() # reads $TESSERA_API_KEY

rel = client.to_duckdb(
    "gold_ohlcv_1m",
    ["BTC", "ETH"],
    tessera.MonthSpan("2026-03", "2026-05"),
)

rel.query("ohlcv", """
    SELECT coin,
           date_trunc('day', time)   AS day,
           last(close ORDER BY time) AS close,
           sum(cvd)                  AS daily_cvd
    FROM ohlcv
    GROUP BY coin, day
    ORDER BY day
""").show()

Prefer raw HTTP? Every partition is also a one-shot Parquet download over the REST API.

Pricing

Start free — including a taste of the funding & volatility forecast. Upgrade for every coin, the full forecast, and the wallet-toxicity factors.

Free

$0 /mo

Get started with no card.

Order-flow-enriched minute OHLCV
A taste of the funding & volatility forecast
BTC, ETH, SOL & HYPE
REST API + Parquet downloads

Pro Recommended

$50 /mo

Everything in Free, plus the alpha datasets.

Everything in Free
The full funding & volatility forecast — every coin
Wallet-behaviour (toxicity) factors
Funding & positioning data, all coins incl. HIP-3

Every plan is all-you-can-eat — unlimited downloads, no per-GB fees. Pull the whole history as often as you like.

Compare plans →