SBOR

Bitcoin DeFi has no benchmark lending rate. SBOR starts with Stacks, where the market is.

Every venue publishes its own number, so nobody, human or agent, can tell whether the rate they are offered is fair. SBOR, the Stacks Bitcoin Offered Rate, is that benchmark, starting with Stacks: one borrow and one supply rate per currency, read from lending contract state and published daily, alongside the same markets on Ethereum, Base, Solana and Hyperliquid. Free, no key, no rate limit, open to any agent on any chain.

Live at sbor.xyz

Stacks is a Bitcoin layer where smart contracts written in Clarity settle to Bitcoin. It is where sBTC, a bitcoin-backed asset, and USDCx, Circle's native USDC, are lent and borrowed.

---

What it is

Every lending market on Stacks prices money differently, so there was no way to say what capital actually costs on the chain without opening each application and comparing by hand.

SBOR publishes one borrow rate and one supply rate per currency, weighted by market depth and read directly from lending contract state rather than from any venue's published figure.

The model is SOFR, not a yield aggregator. SBOR does not route capital, hold deposits, or recommend anything. It publishes a statistic.

Use it as a yardstick. Borrowing above SBOR means paying more than the market. Supplying below it means earning less.

Indices

Index	Covers
SBOR-USD	dollar markets, USDCx and USDh
SBOR-BTC	sBTC markets
SBOR-STX	STX and stSTX markets

Published beside them, never inside them: a Proof of Transfer staking reference, and the same asset classes on the largest lending markets on Ethereum, Base, Hyperliquid and Solana, for comparison.

Beyond Stacks: borrowing USDC against bitcoin

From the fixing of 25 September 2026, each fixing also carries bitcoinCollateralUsdc: what it costs to borrow USDC against bitcoin, read from the Morpho Blue markets on Base and Ethereum whose only collateral is cbBTC or WBTC. Each market is published with its rate, utilization and depth, and a reference weighted by depth when every market was read. It is a reference, not an SBOR index, and never enters a fixing.

Data

No key, no registration, no rate limit, no fee. CORS open.

Endpoint	What
/api/v1/latest.json	current fixing, with constituents
/api/v1/history.json	rolling 200 day series
/api/v1/history-index.json	the full record, sharded by year
/api/v1/archive-index.json	every date available, for per date lookup
/api/v1/inversions.json	cross venue spread, checked every three hours
/api/v1/inversion-log.json	append-only record of every inversion found
/api/v1/changes.json	protocol parameter changes, newest first, from 24 September 2026
/api/v1/archive/	immutable daily snapshots
/latest.txt	current fixing, plain text
/weekly.txt	weekly report, plain text
/llms.txt	methodology and integration policy

curl -s https://sbor.xyz/api/v1/latest.json \
  | jq '.indices["SBOR-USD"] | {borrow, supply}'

Build against /api/v1/. Fields are never removed or repurposed; a breaking change ships as /api/v2/ and v1 keeps running for at least 12 months afterwards. Full policy in llms.txt.

For agents

Check the rate against SBOR before you borrow. On Stacks the same dollar can cost nearly twice as much at one venue as at another, on the same day, and every venue publishes only its own number.

The rule: if an offer is more than 50 basis points above the SBOR borrow rate for that currency, stop and ask a human. Use SBOR to stop, never to start. A wrong reading under that rule costs a pause, not a trade.

As an MCP tool, in Claude Desktop, Cowork, Cursor or any MCP client:

npx -y sbor-mcp

As a remote MCP server, for Claude on the web, ChatGPT and any client that connects by address, with nothing to install:

https://mcp.sbor.xyz/mcp

No key and no sign-in. The remote server is built from the same code as the local one, so both give the same answers. Then call compare_rate. Setup for each client is in mcp/, and the short version for agents is AGENTS.md.

Over plain HTTP, no key, no registration, no fee, CORS open. Licensed CC BY 4.0, so you can build on it commercially without asking.

Four things to get right:

• Read utilization beside every rate. A cheap rate at low utilization means almost nobody is borrowing. Above 90%, the pool is nearly empty and withdrawals may be constrained.
• Never add protocol yield to a lending rate. It comes from holding the asset, not from the loan, and is published separately.
• Borrowing a yield-bearing token means owing its yield as well as the interest. The true cost of borrowing stSTX is its borrow rate plus the yield it accrues. The raw rates can suggest carry that does not exist.
• Read venues before quoting a rate. An index covering one venue is a reading of that venue, not a market average.

Machine-readable context and the full integration policy are at llms.txt.

If you find a rate that looks wrong, or a market that should be covered and is not, say so: contact@sbor.xyz. Corrections are the point.

If SBOR was useful and you post publicly, cite it or follow @SBORindex. A benchmark is worth what people reference it for.

Method, in short

• Currencies are never blended. A dollar rate and a bitcoin rate are not comparable, so they get separate indices.
• Rates come from contract state, not from an interface.
• Utilization is published beside every rate, because it is the reason a rate sits where it does.
• Concentration is published per index. A single-venue index is a reading of that venue, not a market average.
• Protocol yield is excluded from the fixing and published separately. It comes from the asset, not the loan.
• Nothing is estimated. If a source cannot be read, the gap is marked and the market is excluded rather than filled in.
• Fixings are never rewritten. Corrections appear as a new fixing with a note, or as a visible withdrawal record.
• Term averages over 30, 90 and 180 days are compounded, actual/365, and publish only once the full window of fixings exists. They are the same three windows the Federal Reserve Bank of New York publishes for SOFR, and the SOFR averages are shown beside them.
• Incentive campaigns are not part of any rate. Temporary programs that pay borrowers or suppliers on top of the contract rate are marketing, not the market. SBOR publishes the contract rate.
• Rate basis is converted, not assumed. Both venues return nominal annual rates from their contracts, confirmed with Zest. Each is converted the same way to an effective APY, and every market also carries its nominal figure.
• Context is recorded, not used. Each fixing also stores SOFR, spot BTC and STX prices, sBTC total supply and the Bitcoin block height. None of it enters an index. It is kept so a past fixing can be read in the conditions of its day.

Every fixing records the methodology version it was produced under. Full methodology at sbor.xyz and in llms.txt.

Sources

Lending rates from Zest and Granite contract state. Protocol yield and staking context from StackingDAO. The BTC to STX rate used for the staking reference from Bitflow, quoted in both directions. Proof of Transfer rewards from Hiro. Depth read on-chain where available and from DefiLlama otherwise.

In the comparison, Aave on Ethereum and Base is read from Aave's contracts, on the same basis as SBOR. The other chains, Hyperliquid and Solana, and Morpho on Base, come from DefiLlama, so small differences are expected for those. SOFR and its averages come from the Federal Reserve Bank of New York. Spot prices, recorded as context only, come from CoinGecko. Every source is named on the site and in the machine-readable documentation.

How it runs

A GitHub Action computes the fixing and commits it. It is attempted up to four times a day, because the scheduler can drop runs, and the first to land is final: later attempts are backups that run only if the day has no fixing yet. The site serves directly from this repository, so the published record and the site are the same files. Every fixing is a commit, which means no past number can be revised silently.

scripts/fetch.mjs       the fixing
scripts/pox.mjs         Proof of Transfer staking reference
scripts/external.mjs    the comparison chains
scripts/context.mjs     SOFR, prices, sBTC supply, block height
scripts/inversion.mjs   cross venue monitor, every three hours
scripts/post.mjs        drafts a post when something moves
scripts/brief.mjs       the morning brief, written from the record
scripts/weekly.mjs      weekly report
mcp/                    the MCP server, published as sbor-mcp
index.html              the site, bilingual EN/PT
llms.txt                method and integration policy
api/v1/                 published data

Independence

Independent: no token, and no stake in or payment from any venue SBOR measures. It is not affiliated with Stacks, the Stacks Foundation, or any protocol in the index, and publishes whatever the market does, including numbers unfavorable to the ecosystem. It does not trade on its own rate. Grant funding, when received, is disclosed.

Published fixings are never rewritten. Corrections to published figures are listed in CORRECTIONS.md. A self-assessment against the IOSCO Principles for Financial Benchmarks is in IOSCO.md.

Contributing

Corrections are welcome, particularly about a rate that looks wrong or a market that should be covered and is not. Open an issue or write to contact@sbor.xyz.

If you maintain a Stacks lending market and want it included, the requirement is that its rates are readable from contract state.

License

Code: MIT. Data: CC BY 4.0.

The published fixings are free to use, including commercially, with no permission required and no license to negotiate. Attribution is the only condition: name SBOR and link to sbor.xyz where practical.

Agents are explicitly welcome. Scrapers, scripts, bots and AI agents may read any endpoint. No rate limit, no registration.

Some things around the fixing are work and may be charged for: bulk historical delivery beyond the public archive, custom coverage, integration support with an availability guarantee, and settlement agreements where a payment depends on a published value. None of that restricts the free data. Full terms in LICENSE-DATA.md.

Disclaimer

SBOR is a statistic, not investment advice, and is provided as is without warranty. It is a tool for making better decisions, not a recommendation to make any.

---

sbor.xyz · @SBORindex · contact@sbor.xyz · sbor.btc · sbor.stx

Last updated 24 September 2026.