Are you an LLM? Read llms.txt for a summary of the docs, or llms-full.txt for the full context.
Skip to content

Pricing

oculr is metered: you pay for the actual cost of analysing your transaction. No subscriptions, no tiers, no API key required. API and agent requests settle over MPP in USDC.e on Tempo; in the web app you pay from a prepaid balance you fund by card or with USDC.

What you pay

You're charged for exactly the work your transaction needs - LLM reasoning, trace fetches, address labels, token prices, web research, code execution. A simple transfer touches a handful of upstream services; a deep exploit investigation runs many more tool calls and burns far more tokens, so it costs more.

Charges settle as the analysis runs, each one the exact amount owed since the last - you never prepay for work that didn't happen, and there's no flat quote to overshoot.

The two endpoints bill at different granularity:

  • Sync POST /explain (SSE) - metered exactly. You pay your specific analysis's cost, signed incrementally as it runs.
  • Async POST /explain/async - collected across the job lifecycle. Submitting charges $0.01; each GET /result/:jobId poll collects what the analysis has accrued since your previous poll; the first poll after the job finishes charges the true-up. The cumulative total is identical to the sync SSE price for the same analysis. A poll with nothing yet to collect is free, as are subsequent fetches of an already-paid finished result. Failed jobs true up the same way. Keep polling with a paying client: 90 seconds with nothing collected aborts the run and leaves a partial result.

Paying in the web app

The web app charges the same metered price, deducted from a prepaid balance as the analysis runs. You fund that balance one of two ways:

  • By card. Sign in with Google and buy credit on Stripe's secure checkout page: any amount from $10. $10 buys $10 of credit. Unused credit can be refunded to your card on request. A Google account's first 10 analyses are free (limited to a daily total across all users).
  • With USDC. Connect a wallet, sign in, and send USDC on Tempo to your personal deposit address. Unspent balance can be withdrawn to the wallet you signed in with.

The two are separate accounts with separate balances. Opening a past analysis from My reports is free. See Use the web app for the steps.

What analyses cost

Most analyses cost between $1 and $2:

  • Routine work - transfers, swaps, straightforward DeFi - runs about $1.
  • Incident and exploit investigations - many more tool calls over far more data - usually $1.50 to $2.50, with the occasional deep investigation reaching about $5.

Prices fell sharply at the end of September 2026. The figures previously published here (a $3.68 median and an $8.06 90th percentile, measured to 2026-07-29) no longer apply.

For reference, the distribution across 67 production analyses from 2026-09-19 to 2026-10-19 on claude-opus-5-5 (the default model), at the contract price:

Charge
Cheapest$0.11
25th percentile$0.91
Median$1.27
75th percentile$1.67
90th percentile$2.24

Size your deposit against the top of the range, not the median. One analysis in ten costs more than $2.24, and the most expensive analysis in the window was $4.97. The suggested deposit of $16 covers that several times over.

Two honest caveats about that table. It covers every production analysis in the window, oculr's own benchmark and verification runs included, so it is not a picture of customer traffic alone: 67 analyses over 47 distinct transactions, 18 of them on the benchmark corpus, which is deliberately weighted toward hard exploit cases because those are what the service is built for. And it is a small sample - it describes what we have measured, not a guaranteed bound.

What kind of transaction it is

What kind of transaction you send explains much of that spread, though not all of it. Joining production charges to the benchmark corpus by transaction hash prices each class of work directly:

Transaction typeAnalysesDistinct txsTypical chargeRange
Benign control (transfer, swap, liquidation, bridge deposit)64$0.92$0.72 - $1.00
Signature replay11too few to say$0.70 - $0.70
Bridge / cross-chain message validation31too few to say$0.65 - $1.27
Admin key compromise21too few to say$0.83 - $2.02
Reentrancy11too few to say$1.61 - $1.61
Access control bypass21too few to say$1.67 - $1.87
Oracle / price manipulation22too few to say$1.67 - $1.92
Rounding / share inflation11too few to say$2.24 - $2.24

Read the two count columns before the prices. Analyses is how many paid runs went into the cell; distinct txs is how many different transactions those runs covered. Three runs over one transaction is one transaction priced three times, not a market rate - so the thin rows quote no typical figure at all rather than dressing up a median over three runs. Even the fullest cell rests on four transactions.

What the table does support:

  • Benign transactions are the cheapest class by median. Four ordinary mainnet transactions - a DEX aggregator swap, an Aave liquidation, a bridge deposit, an NFT mint - at a median of $0.92 over 6 runs, none above $1.00. That is the routine band, measured rather than assumed.
  • An exploit still costs more than a control, but the gap has narrowed. Across the 12 exploit analyses the median is $1.61, against $0.92 for the 6 controls, and no exploit run in the window exceeded $2.24.
  • Every exploit class is thin - one or two transactions each - so no exploit class quotes a typical figure. Read the range column as the measurement and the class as a label, not a price list.
  • The spread inside a class is real, and it is not something you can plan around. The same TesseraDAO transaction billed $0.83 and $2.02 on two runs. How deep an investigation goes varies run to run, so treat a class median as a centre of gravity, never as a quote.

Same population and method as the distribution above: production runs on claude-opus-5-5, at the contract price, 2026-09-19 to 2026-10-19, deduped by request. 18 of the 67 production analyses in that window ran on a corpus transaction and are in this table; the other 49 ran on transactions outside the benchmark corpus, with no attack type to report.

Regenerate rather than edit: npx tsx --env-file=.env scripts/pricing-by-attack-type.ts.

Choosing a cheaper model

Model reasoning is roughly half the bill - measured at 48% of the charge against 51% for data collection, across the 15 recorded production runs. The rest (traces, labels, prices) does not change with the model, so a cheaper model cuts at most about half your cost, not all of it. Pass a model field on the request body to trade accuracy for cost:

ModelRelative LLM costWhen to use
claude-opus-5-5 (default)1×Best accuracy - incident triage, exploits, anything you'll act on
claude-opus-51.25×Prior Opus generation
claude-opus-4-81.25×Prior Opus generation
claude-opus-4-71.25×Prior Opus generation
claude-sonnet-5~0.5×Routine DeFi decoding at volume
claude-haiku-4-5-20251001~0.25×Bulk classification, simple transfers

Those ratios are per-token input/output list prices ($4/$20 per Mtok for Opus 5.5, $5/$25 for earlier Opus, $2/$10 for Sonnet, $1/$5 for Haiku). Cached context is a large share of every run and does not follow the same ratios: it is $0.20/Mtok on both Opus 5.5 and Sonnet, against $0.50 on earlier Opus. So earlier Opus costs more than 1.25× in practice, and Sonnet saves less than the ratio suggests. Applied to a ~48% LLM share, Sonnet saves at most roughly 25% of the total bill and Haiku at most roughly 35%.

This list is the server's actual allowlist (ALLOWED_MODELS); any other value is rejected. GET /tool-spec.json publishes the live default in its defaultModel field if you'd rather read it programmatically than trust this page.

How metering works

Sync analyses use MPP metered sessions - the streamed-payments variant of the protocol's session intent:

  1. Your client opens a payment channel against the API with a maxDeposit cap.
  2. As the analysis accrues cost, the server requests a voucher for the exact amount owed so far, so your cumulative payment tracks the running cost. The session client signs them automatically mid-stream - no interaction.
  3. The final increment settles the balance, and the result is released.

Two properties fall out of this design:

  • You never prepay for work that didn't happen. If an analysis is cheap, you pay a cheap price. There is no flat quote to overshoot.
  • Your hard spend cap is maxDeposit. The server can never charge past the channel deposit, and your client will never sign past maxDeposit. It is a ceiling, not the amount you escrow - so give it real headroom above your most expensive expected analysis, since an exploit investigation can run to several dollars. See The channel deposit for why it must sit above the suggested deposit rather than on it.

The channel deposit

Opening a channel escrows a deposit. Every 402 from oculr advertises a suggestedDeposit of $16, and your client opens at min(suggestedDeposit, maxDeposit).

The deposit is escrow, not a charge. You are still billed only for the analysis you actually run, and every unspent cent of the deposit returns to your wallet when the channel closes. A $16 deposit on a $1.20 analysis means $1.20 spent and $14.80 refunded, not $16 spent.

Size it above your analysis, not below. Once cumulative spend reaches the deposit, your client has to raise it on-chain, mid-analysis, and mppx sizes that top-up to the $16 suggestion it was offered on the challenge, bounded by the room left under your maxDeposit - so a deposit that runs out stops the stream for one on-chain top-up at the recommended '32' cap rather than for the amount it was actually short (a larger cap leaves room for a further $16 step at each later crossing). $16 clears every production analysis on record, but the hardest cases come within a few dollars of it - treat $16 as the floor for a session that might hit a hard case, not as generous headroom.

Set maxDeposit above the suggestion, not equal to it. maxDeposit is a ceiling, not escrow - the channel still opens at $16 (the smaller of the two), so a higher cap costs you nothing up front. But set it on $16 and the channel can never grow: mppx won't sign past the ceiling, so a session that later needs more - a second analysis on the same channel, or a hard case near the top of the range - fails mid-run with no result. Use '32' - it escrows the same $16, with headroom for the hardest cases and a second analysis on one channel.

The headroom is a fallback, not a budget. It is there so the automatic top-up has somewhere to go if a run does overrun, which is the only route by which an overrunning run finishes at all - degraded, but not the hard stop an equal cap gives. There is no working out-of-band way to raise a live channel by hand: session.topUp() signs against the challenge minted for the first request on the session, and challenges expire after 30 minutes, so a session that outlives one gets a 402. Closing and reopening does not consume the headroom either - close() resets cumulative spend to 0, so the fresh channel opens at min(suggestedDeposit, maxDeposit) = $16 whatever the cap. So don't let one run drift past its deposit into an on-chain top-up mid-stream.

import { tempo } from 'mppx/client'
import { privateKeyToAccount } from 'viem/accounts'
 
const account = privateKeyToAccount(process.env.WALLET_PRIVATE_KEY as `0x${string}`)
// Two different numbers, on purpose. The channel opens at
// min(suggestedDeposit, maxDeposit) = $16 - that is what gets escrowed, and
// whatever you don't spend is refunded on close. maxDeposit is the signing
// ceiling: '32' leaves the channel room for the automatic top-up. Equal to the
// suggestion it could never grow at all, and the second heavy run would die
// mid-stream.
const session = tempo.session.manager({ account, maxDeposit: '32' })
// session.sse('https://mpp.oculr.xyz/explain', …) - see the quickstart for the full call.

Because metering rides the stream, incremental vouchers apply to the SSE endpoint. The async endpoint reaches the same total through the classic Mppx.create() + fetch() flow: $0.01 at submit, then one charge per poll for whatever accrued since the previous one, and a true-up on the first poll after the job finishes. Full recipes for both: Call the oculr MPP.

Cost transparency

The price is on the wire, not in the result body. On the SSE endpoint your session client sees every increment it signs as the metered stream charges; their sum is what the analysis cost. On the async endpoint it is the amount on each 402 challenge, settled as described above. Reconcile against those figures: the result JSON carries no cost breakdown.

Watching the live stream? SSE mode emits progress events (tool_call, tool_result, tokens, …) as the analysis runs. See Core concepts → SSE streaming.

What's free

Discovery and health surfaces cost nothing - no payment challenge, no wallet needed:

PathWhat it is
GET /healthService health probe
GET /openapi.jsonFull OpenAPI 3.1 spec
GET /tool-spec.jsonTyped Anthropic + OpenAI tool schemas
GET /SKILL.mdProse entry point for agents
GET /llms.txtDiscovery index for LLM crawlers

FAQ

Is there a rate limit? No fixed limit - per-request payment is the throttle, and maxDeposit is your cap.

What if the analysis fails? oculr returns partial results (HTTP 200, confidence: 'low', a summary that starts with "Partial result") rather than failing outright - you're only metered for the work that ran. See Partial results.

Where does my payment go? API and agent payments settle in USDC.e on Tempo, token contract 0x20C000000000000000000000b9537d11c60E8b50; funding instructions are in the FAQ. Card top-ups in the web app are processed by Stripe; oculr never sees your card details.

Related

  • Call the oculr MPP - mppx setup and your first paid request
  • Core concepts - sessions, confidence levels, partial results
  • Use the web app - signing in with Google or a wallet, card and USDC top-ups
  • FAQ - wallet funding, supported chains, troubleshooting